Content uploaded by Qian Yang
Author content
All content in this area was uploaded by Qian Yang on Feb 23, 2024
Content may be subject to copyright.
The Future of HCI-Policy Collaboration
Qian Yang
Cornell University
USA
qianyang@cornell.edu
Richmond Y. Wong
Georgia Institute of
Technology
USA
rwong34@gatech.edu
Steven J. Jackson
Cornell University
USA
sjj54@cornell.edu
Sabine Junginger
The Lucerne University of
Applied Sciences and Arts
Switzerland
sabine.junginger@hslu.ch
Margaret Hagan
Stanford University
USA
mdhagan@stanford.edu
Thomas Krendl Gilbert
New York Academy of
Sciences
USA
tgilbert@nyas.org
John Zimmerman
Carnegie Mellon University
USA
johnz@cs.cmu.edu
ABSTRACT
Policies signicantly shape computation’s societal impact, a crucial
HCI concern. However, challenges persist when HCI professionals
attempt to integrate policy into their work or aect policy outcomes.
Prior research considered these challenges at the “border” of HCI
and policy. This paper asks: What if HCI considers policy integral to
its intellectual concerns, placing system-people-policy interaction
not at the border but nearer the center of HCI research, practice,
and education? What if HCI fosters a mosaic of methods and knowl-
edge contributions that blend system, human, and policy expertise
in various ways, just like HCI has done with blending system and
human expertise? We present this re-imagined HCI-policy rela-
tionship as a provocation and highlight its usefulness: It spotlights
previously overlooked system-people-policy interaction work in
HCI. It unveils new opportunities for HCI’s futuring, empirical,
and design projects. It allows HCI to coordinate its diverse policy
engagements, enhancing its collective impact on policy outcomes.
CCS CONCEPTS
•Human-centered computing
→
Empirical studies in HCI;
HCI theory, concepts and models;•Computing methodologies
→
Articial intelligence.
KEYWORDS
Policy, design, societal impact of technology.
ACM Reference Format:
Qian Yang, Richmond Y. Wong, Steven J. Jackson, Sabine Junginger, Margaret
Hagan, Thomas Krendl Gilbert, and John Zimmerman. 2024. The Future of
HCI-Policy Collaboration. In Proceedings of the CHI Conference on Human
Factors in Computing Systems (CHI ’24), May 11–16, 2024, Honolulu, HI, USA.
ACM, New York, NY, USA, 15 pages. https://doi.org/10.1145/3613904.3642771
Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for prot or commercial advantage and that copies bear this notice and the full citation
on the rst page. Copyrights for components of this work owned by others than the
author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or
republish, to post on servers or to redistribute to lists, requires prior specic permission
and/or a fee. Request permissions from permissions@acm.org.
CHI ’24, May 11–16, 2024, Honolulu, HI, USA
©2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.
ACM ISBN 979-8-4007-0330-0/24/05. . . $15.00
https://doi.org/10.1145/3613904.3642771
1 INTRODUCTION
From the future of work, to equity and sustainability, computation’s
societal impact has become a critical HCI concern. At least three
forces help shape this impact: the incentives, deterrents, and rules
wired into computational systems (“system design”); the social and
cultural tendencies of people (“social practice”); and the laws and
regulations that govern both systems and people (“policy”) [
20
,
43
,
47
,
96
]. Take privacy as an example: The design of a mobile app,
its users’ privacy awareness, and related policies (e.g., app stores’
mandates, governments’ regulatory nes) all help decide how well
this app will protect user privacy [42].
In many cases, HCI researchers and practitioners are already suc-
cessful in improving system-people-policy interaction. For instance,
HCI scholar Lorrie Cranor both conducted empirical and design re-
search on privacy, and led the drafting of privacy regulations at the
U.S. Federal Trade Commission (FTC) [
24
]. Professor Colin Gray’s
work on deceptive web design patterns inuenced the U.S. DETOUR
(Deceptive Experiences To Online Users Reduction) Act [
5
,
36
].
Such success stories span many countries and social issues, such
as content moderation [
39
], gig worker protection [
61
,
67
,
85
], AI
safety [87], smart cities [26, 91], sustainability [14], and more.
Yet in many other cases, policy considerations lag behind, or
remain isolated from) the other HCI work that centers around
systems and people. Consider Uber’s creation of gig work (and
intentional breach of taxi regulations) when it rst launched its
ride-sharing service [
12
,
51
], Facebook’s adoption of AI newsfeed
rankers (and amplication of misinformation) [
39
], or OpenAI’s
public release of chatGPT (and intentional violation of copyright
laws) [
83
,
92
]. Too often, HCI seemed to become excited about the
technology rst, take an interest in policy only after things had
gone o the rails, and then nd itself in the unenviable position of
trying to put genies back into bottles.
This paper aims to better understand this disparity and amplify
HCI’s collective voice in the policy realm. From diculties in bring-
ing policymakers to the table [
77
], to misalignment between HCI
methods and policymakers’ evidentiary needs [
78
], many chal-
lenges of HCI-policy collaboration are well-known. Prior work
referred to these challenges as at the “boundaries” of HCI and pol-
icy [
77
], prompting decade-long calls for more collaborations across
HCI-policy disciplinary boundaries [
20
,
24
,
43
,
47
,
57
,
58
,
86
]. These
calls are necessary and valuable [
60
]; nothing in this paper argues
CHI ’24, May 11–16, 2024, Honolulu, HI, USA Yang et al.
against them. Meanwhile, the decade-long persistence of these calls
also raises the question: Can fresh perspectives on these challenges
inspire additional strategies for overcoming them? This paper aims
to provide one such perspective.
This paper argues that—thanks to the decade-long calls for more
HCI-policy collaboration—HCI has begun creating new methods
and new types of knowledge contributions that are distinct from
traditional HCI or policy ones. HCI and policy expertise have begun
cross-pollinating. However, such unconventional, trans-disciplinary
work has seemed to struggle with getting past reviewers, therefore
has remained largely unrecognized in peer-reviewed HCI research
venues, industry practices, or HCI textbooks. This lack of recog-
nition can contribute to the disparity in HCI’s policy engagement
and impact.
Reframing the challenges of HCI-policy collaboration as “in-
sucient recognition of trans-disciplinary methods and knowledge”
informs new strategies for addressing them: What if HCI were to see
policy as integral to its intellectual concerns, placing system-people-
policy interactions not “at the boundaries of HCI and policy” [
77
], but
nearer the center of HCI research, practice, and education? What
if HCI were to foster a mosaic of distinctively HCI methods and
knowledge contributions that blended system, human, and policy
expertise in varying ways and degrees, just like it has previously
done with blending system and human expertise?
This paper sketches out this re-imagined HCI-policy landscape
as a provocation. We demonstrate how it can (1) spotlight previously
overlooked system-people-policy interaction work and knowledge,
(2) unveil new opportunities for HCI’s futuring, empirical, and de-
sign projects, and (3) allow HCI to coordinate its diverse policy
engagements, thereby enhancing HCI’s collective impact on policy
outcomes. We show how it opens up many new research questions,
therefore is a useful seed for broader, HCI-community-wide discus-
sions. We invite fellow HCI researchers and practitioners to discuss
their views of the relationship among systems, people, and policy
expertise, and how HCI wants to position itself in this relationship.
This paper makes three contributions. First, it moves the research
discourse on HCI-policy collaboration beyond the decade-long calls
for more collaboration, towards a community-wide discussion on its
problem-solution framing. Second, it jump-starts this discussion by
proposing one new problem-solution frame: There is insucient
recognition that policy is not just HCI’s “broader impact”, but in-
tegral to HCI’s intellectual pursuits in system-people interaction.
Finally, it exposes important new questions for future research to
address. For example, when is policy not the best solution to tech’s
societal harms? And when sudden momentum for policy change
appears, how might HCI quickly assemble its diverse work and
seize the opportunity?
2 KEY CONCEPTS
At the outset of this paper, we plunged head rst into the topics of
“policy” and “HCI-policy collaboration” without detailing what they
meant. These are nuanced concepts, variably interpreted and lack-
ing agreed-upon denitions [
49
,
68
,
89
]. With these complexities in
mind, we unpack what we mean by these terms. Our goal is not to
establish conclusive denitions, but to scope them for the narrow
purpose of this paper.
2.1 Policy
In this paper, policy refers to principles, guidelines, and written
rules formulated and adopted by an authority, organization, or
government [
43
]. Among the widely-accepted policy denitions,
we chose this one because it has both the specicity and breadth
that match our research goals.
First, it encompasses various rules external to a computing sys-
tem that wield authority over human-computer interactions, such
as governmental policies, tech platform policies, and more. All these
policies help shape computation’s societal impact, therefore hold
relevance to HCI.
Secondly, this denition recognizes that a policy encompasses
not only its written rules, but also the ways in which these rules are
formulated and applied. Take the EU AI Act as an example. Whose
inputs shaped this legal framework [
69
], how government agencies
translate it into enforceable rules, and how the courts interpret and
apply the rules to specic situations [
90
] all inuence AI’s societal
impact. All are relevant to HCI.
This denition of policy (and the scope of this paper) is also
deliberately narrower than some prior work. It does not intend to
address all aspects of policy practices, nor all political forces that
can inuence computation’s societal impact. Because our goal is to
understand the challenges of HCI-policy collaboration, we focus on
the aspects of policy uncommon in current HCI work. For example,
a society’s cultural politics (e.g., its conception of fairness [
63
])
and governmental policy-making procedures both signicantly
inuence its tech policies. Yet because the former is already common
in HCI work, the latter takes precedence in this paper.
2.2 “Understanding” Policy, “Designing” Policy
What this paper refers to as understanding and designing policy
reects this policy denition. Because "policy" encompasses both
the written rules and the process of their formulation and enact-
ment, “understanding" a policy includes understanding both aspects.
Similarly, “designing” policy includes not only formulating rules,
but also designing the processes for formulating and implementing
the rules [
71
]. In practice, policy design is the iterative process of
(1) identifying policy needs, (2) clarifying policy needs (or issue-
framing), (3) formulating policy, (4) designing systems and services
that implement policy, and (5) evaluating policy outcomes [
49
,
70
]
1
.
We chose a broader view of designing policy because we want to
consider various HCI-policy connections [28, 94].
The synergies between technology and policy design are self-
evident. Both processes start with recognizing an opportunity
for new systems/policies to help, and proceed with prototyping
and evaluating their helpfulness iteratively. These synergies have
prompted some HCI researchers to call for designing technology
and policy simultaneously [
96
] and some legal scholars to call HCI
a new direction in designing Information Technology law [86].
1
This scope of policy “design” is intentionally broader than some prior HCI liter-
ature, which focused on evidence-based policy-making (and HCI as a supplier of
evidence) [
77
]. It is also broader than some legal scholarship, which sees computing
system “design” as merely the implementation of a given policy requirement (e.g.,
generating a standard cookie opt-out menus [
84
], rather than a process of open-ended
knowledge inquiry.
The Future of HCI-Policy Collaboration CHI ’24, May 11–16, 2024, Honolulu, HI, USA
2.3 “Productive” HCI-Policy Collaboration
By denition, HCI addresses the interactions between comput-
ing systems and people. As these systems grew ubiquitous and
integral to everyday life, HCI has expanded its focus from single-
user-single-system interactions, to interactions among platforms,
stakeholders, organizations, societal processes, and even planetary
concerns [
14
,
32
,
95
]. Policy entered HCI discourse during this ex-
pansion, because it signicantly shapes these larger-scale and often
political interactions [27, 43, 60].
A productive HCI-policy collaboration can improve computa-
tion’s societal impact, by catalyzing synergistic designs of computa-
tional systems, social practices, and policies [
29
,
43
]. Consider, for
example, a novel AI system that helps diagnose a disease (system),
clinician teams’ diagnostic workow (social practice), and clinical
malpractice laws (policy). To improve diagnostic accuracy in prac-
tice, designers cannot simplistically match clinician workows and
malpractice laws with this new AI system, nor vice versa. Instead,
the three designs need to co-evolve [
70
]. Computation’s societal
impact improves step-wise, as the “knot” of system design, social
practice design, and policy design loosens and tightens, unwinds,
and reties [43].
There is no single best approach to HCI-policy collaboration that
prescribes this ideal outcome [
30
]. Should HCI professionals gen-
erate empirical evidence of technological harm, before persuading
policymakers to act [
37
]? Or should they envision a computational
system and its regulations simultaneously [
96
]? Should they seek a
government position or work with advocacy groups when prompt-
ing new policies? Depending on context, the most fruitful approach
to HCI-policy collaboration varies, each bringing dierent chal-
lenges.
2.4 Problem-Solution Frames
How might we grasp the challenges of HCI-policy collaboration and
devise solutions? This is a wicked problem [
73
]. Because challenges
facing every collaboration vary, there is no denitive answer to the
question of “which challenges(s) are responsible for the disparity in
their success levels overall” or “what solutions might help” [27].
Addressing wicked problems relies on problem framing and
reframing [
73
]. By seeing a problematic situation through dierent
lenses, one can better understand its various dimensions and see
new avenues for potential solutions [
28
]. And this is the goal of
this paper. We wanted to nd a new frame for understanding the
challenges and disparity in HCI-policy collaboration; a frame that
can reveal new avenues for solutions.
3 BACKGROUND
HCI-policy collaboration has many well-known success stories and
persistent challenges. For example: diculties in bringing policy-
makers to the table [
77
], policymakers demanding policy evidence
that (1) may not exist (e.g., evidence of a new technology’s not-yet-
manifested societal form) or (2) may not be the types of evidence
HCI approaches produce [
77
,
78
]; mismatch between the pace of
policy or political changes and the pace of HCI work [24, 66]; and
more.
Interestingly, prior research rarely deliberated on the nature of
these problematic situations. Instead, it directly suggested solutions.
Most commonly, researchers called for more HCI-policy collabo-
ration [
20
,
24
,
43
,
47
,
57
,
58
,
86
]. Others suggested more concrete
solutions. For instance, to bring policymakers to the table, they
encouraged more HCI scholars to take a gap year and work at
government agencies [
59
]. To address the HCI-policy evidentiary
gap, they recommended involving policymakers in participatory
design workshops, to help them appreciate HCI methods [
77
]. To
align the misaligned timelines of policy and HCI work, HCI re-
searchers recommended committing to policy work for the long run
and adjusting to the inherently slower pace of policy change [
24
].
These recommendations yielded valuable results [
60
]; nothing in
this paper argues against them. Nevertheless, we see opportunities
to deliberate on the problem-solution frames that these solutions
imply. Let us illustrate these opportunities through two examples.
3.1 Current Problem-Solution Frame #1
The numerous calls for more HCI-policy collaboration suggest
that a critical hindrance to this collaboration is that too few HCI
professionals are participating [57].
However, the situation might be shifting. Over the past decade,
methods such as participatory design and value-sensitive design—
methods many policy actors also use—have moved more towards
the center of HCI [
33
,
74
]. Since 2018, the ACM FAccT conference
has been bringing together HCI, law and policy, and other elds
to address AI’s ethical issues [
56
]. Since 2021, SIGCHI publications
mentioning “policy” surged by more than 40%, according to ACM
Digital Library In this context, it is worth asking whether HCI’s
lack of attempts to policy engagement remains true today.
3.2 Current Problem-Solution Frame #2
HCI researchers have characterized the challenges of HCI-policy
collaboration as occurring at the boundaries of HCI and policy [
77
].
This framing sees policy change as a “broader impact” of HCI’s
intellectual work (i.e., understanding and designing system-people
interaction). Therefore, the disciplinary boundaries between HCI
and policy can hinder their collaboration.
Along this line, prior research repeatedly recommended that HCI
and policy communities accept each other’s norms. For example,
methodologically, HCI researchers recommended that policymakers
embrace HCI’s design methods [
77
]; socially, they encouraged more
HCI professionals to socialize with policy actors [
57
,
59
]; temporally,
they encouraged HCI researchers to adapt to policy and political
timelines [24].
These suggestions are highly valuable, but seem to be partial so-
lutions for fostering productive HCI-policy partnerships. Take, for
instance, the timeline of addressing the societal impact of generative
AI (genAI). How should HCI and policy communities coordinate
their pace of work to best address genAI’s societal impact? How
should HCI communities approach publishing novel GPT applica-
tions, considering some might be soon regulated out of existence?
These questions require nuanced debates [
41
]. Urging one commu-
nity to adopt the other’s timeline appears insucient.
More fundamentally, a productive collaboration between any two
disciplines involves more than choosing whose norms to conform to.
It entails creating new trans-disciplinary methods and new bodies
of knowledge distinct from either parent discipline [
44
,
54
]. The
CHI ’24, May 11–16, 2024, Honolulu, HI, USA Yang et al.
innovations of computing systems, social practices, and policies
require distinct methods, and each encompasses multiple temporal
patterns [
64
,
80
,
81
]. Are there trans-disciplinary methods that can
respect and bridge these dierences? Such meta-discussions are
absent in current HCI-policy research.
4 RESEARCH ACTIVITIES
We (a group of HCI, design, law, and policy researchers) wanted to
reframe the challenges facing HCI-policy collaboration, in the hope
of identifying new opportunities for addressing them. Towards
this goal, we collected empirical reports of previous HCI-policy
collaboration processes, analyzed factors contributing to their vary-
ing levels of success, and then worked to identify a new, useful
problem-solution frame.
4.1 Literature Review
We rst searched peer-reviewed HCI research publications and
practitioner-facing books for empirical reports of previous and on-
going collaboration processes. However, we soon realized that this
literature has published very few such reports. Even the most fruit-
ful collaborations—collaborations that resulted in national policies
with HCI’s scholars’ names [
24
,
36
]— left little documentation of
how HCI researchers and practitioners worked with policy actors in
practice, or how they approached the gap between HCI and policy.
HCI’s Participatory Design (PD) projects are telling examples.
Peer-reviewed HCI venues have published many such projects,
where researchers investigated vulnerable stakeholders’ needs and
engaged them in drafting policy recommendations [
91
,
97
]. While
these publications oered detailed need-nding and rule-making
processes, it is dicult to gauge whether or how they inuenced
policy outcomes. After all, citing academic references or attributing
individual researchers is uncommon in many legal or public policy
documents.
A larger set of HCI-policy collaboration eorts reside outside
of peer-reviewed HCI publications. We found their traces in less
prominent genres of HCI research dissemination, e.g., in <Interac-
tions> magazines [
58
,
66
], in non-archival policy white papers [
83
],
in the news [
18
], and on Medium [
50
]. Unfortunately, these publica-
tions also provided little detail about the HCI-policy collaboration
contexts, processes, or outcomes.
Other collaboration eorts seemed entirely untraceable. For ex-
ample, we suspect HCI research has inuenced Facebook’s content
moderation policies, because many HCI scholars worked on the
topic as Facebook employees. However, such collaboration is di-
cult to verify, much less to study rigorously.
4.2 Community Inputs and Discussions
To collect more data on HCI-policy collaboration processes and
breakdowns, we held a workshop on this topic at the 2023 CHI con-
ference. With an open call for participation, the workshop brought
together 57 HCI researchers and practitioners from the Americas,
Asia, Australia, and Europe. The workshop received and published
more than 60 papers detailing the participants’ specic research
projects, all at some intersection of computational systems, peo-
ple, and policy. They cover an overwhelming breath of topics, e.g.,
social media’s role in human tracking, surveillance of migrant
workers, unions negotiating tech policy for workers, combined use
of HCI and policy in tenant protection, policy issues throughout
the supply chain of generative AI, AI in e-governments, challenges
in applying GDPR to user interface design, and many more. These
publications became an additional source of data for our analysis.
4.3 Data Limitations
One limitation of our data is geographical. While workshop partic-
ipants represent many regions globally, all authors of this paper
are based in the U.S. and Europe, as are most workshop publication
authors. All except one workshop author are based in democra-
cies. Recognizing this geographical bias, we strongly encourage
researchers from other parts of the world to help critique this work
and share their perspectives on HCI-policy collaboration.
4.4 Data Synthesis
We synthesized a new problem-solution frame based on this rela-
tively diverse set of empirical reports. As we will demonstrate in the
remainder of the paper, this frame can (1) help explain the observed
disparity in HCI-policy collaboration and (2) reveal new avenues of
opportunity in addressing the disparity. It is (3) exible, allowing
HCI researchers and practitioners to derive specic actions accord-
ing to the respective context of their policy engagement. Finally,
even the authors lack consensus on whether this problem-solution
frame is too progressive and controversial, or too obvious that is
what HCI needs to do. In this sense, this frame can be (4) an eective
seed for broader HCI-community-wide discussion.
This frame results from the authors’ year-long discussion within
our small team, with inputs from wider our respective communities.
Appendix A describes our deliberation process in more detail.
5 POLICY AS INTEGRAL TO
HCI’S INTELLECTUAL CONCERNS
5.1 A Vision of the Future
We envision a future in which HCI sees policy as integral to its
intellectual pursuits, placing system-people-policy interaction not
at the boundaries of HCI and policy, but nearer the center of HCI
research, practice, and education. This vision diers from prior
framings in three important ways.
1.
HCI will deliberately dissolve the boundary between its policy-
and system-human-interaction-focused work. Instead, a wealth
of hybrid methods and hybrid knowledge contributions will
bridge and harmonize HCI’s system, people, and policy exper-
tise.
2.
Each HCI project will choose to integrate system, people, and
policy considerations to varying degrees and in diverse modes.
For example, some projects will advance HCI’s knowledge of
system-human interaction, while providing policy implications.
Some projects will leverage human-system interaction expertise
and advance HCI’s knowledge of policy design. Other projects
will generate new knowledge on system-human-policy inter-
action, by bridging the three areas of expertise. In this future,
peer-reviewed HCI research venues will critically assess and
accept these diverse knowledge contributions. HCI education
and practitioner methods will embrace these varied methods.
The Future of HCI-Policy Collaboration CHI ’24, May 11–16, 2024, Honolulu, HI, USA
3.
HCI communities will actively coordinate these varied meth-
ods and diverse knowledge contributions, maximizing HCI’s
collective impact on real-world policy outcomes.
5.2 A New Problem Frame
There are three key dierences between this vision of the future and
the current landscape of HCI-policy collaboration. These dierences
oer us new insights into why HCI-policy collaboration eorts have
had disparate results.
1) Lack of recognition for the inherent tensions between HCI
and policy methods, leading to a lack of recognition for trans-
disciplinary methods that address the tensions. Peer-reviewed
HCI research venues, practitioner-facing methods, and HCI text-
books did not always give due recognition to unconventional meth-
ods and knowledge that bridged HCI-policy expertise. For example,
while many peer-reviewed CHI and CSCW papers feature policy
recommendations derived from PD workshops (a well-established
HCI method) [
33
,
82
], none featured policy proposals derived from
researchers’ own synthesis (for example, by synthesizing policy-
makers’ evidentiary needs and HCI’s prior empirical work) [
24
].
This may be because HCI has not yet acknowledged synthesis as
an established HCI research method, or policy design as an HCI
knowledge contribution. Consequently, methods that can bridge
HCI-policy evidentiary gap remain largely visible in prominent
HCI venues.
This lack of recognition can contribute to the disparity in HCI-
policy collaboration. It can disincentivize early-career researchers
from participating. It also makes HCI-policy knowledge less accessi-
ble, adding to why socializing with policymakers appears to be the
only way to gain such knowledge [
24
,
59
,
77
]. Both eects privilege
established HCI experts (who are more likely to prioritize long-term
policy impact and relationship-building over near-future publica-
tions). Both privilege issues aligned with policymakers’ existing
interests over precautionary debates.
2) Challenges in deliberating and curating novel HCI-policy
methods and knowledge contributions. The limited visibility
of novel HCI-policy methods and contributions not only hinders
methodological innovation. It also inhibits broader HCI communi-
ties from scrutinizing or deliberating upon these emergent methods
and ndings. For example, while many Speculative Design and
PD projects have discussed their implications for policy, few have
asked: Are “implications for policy" a necessary or good measure
of the quality of empirical work that studies technological harm?
What makes an implication for policy more meaningful or useful
than others?
The lack of deliberation on emerging HCI-policy methods means
fewer proven method choices for HCI professionals attempting to in-
tegrate policy into HCI work. Little publicly available guidance exist
on how to write better “Implications for Policy." HCI practitioners’
toolbox contains few methods for designing system-people-policy
interactions. This lack of public knowledge yet again disadvantages
early career HCI professionals.
3) Insucient community-wide coordination for collective
impact. Eorts to coordinate HCI’s diverse modes of policy en-
gagement are scarce in HCI literature. Considering that policies
underpin every computational system and every human, when
is policy design not the best approach to improve a system’s hu-
man impact? Knowing that momentum for policy change can arise
abruptly, how can HCI quickly mobilize its diverse HCI-policy
work to seize the opportunity? Answers to these meta-questions
can amplify HCI’s collective inuence on policy outcomes, yet are
noticeably absent in today’s HCI research discourse.
With little community-wide coordination and support, HCI re-
searchers and practitioners relied on their respective eorts for
policy impact, exacerbating disparities in HCI-policy collaboration
and weakening the collective impact of HCI on policy.
5.3 A New Solution Frame
These new problem frames reveal new solution frames; new strate-
gies for addressing the disparity in HCI-policy collaborations.
1.
Explicate underlying tensions between HCI and policy methods;
2.
Give due recognition to the trans-disciplinary methods that ef-
fectively address these tensions. Assess and accept their knowl-
edge contributions to peer-reviewed HCI research venues, prac-
titioners’ toolboxes, and HCI textbooks;
3.
Foster a mosaic of trans-disciplinary methods and knowledge
contributions that blend policy and HCI’s futuring, empirical,
and design expertise to varying degrees and in diverse ways.
Each individual HCI project can choose among them;
4.
Coordinate HCI’s diverse policy engagements and maximize
HCI’s collective impact on policy outcomes.
The remainder of the paper demonstrates the usefulness of this
new problem-solution frame. We illustrate how even seeing ex-
isting literature through this new frame can reveal tangible, new
opportunities for individual HCI projects (Chapter §6) and for HCI
communities as a whole (Chapter §7).
Do HCI communities endorse this re-imagined HCI-policy re-
lationship? Are these strategies indeed eective in fostering HCI-
policy collaboration and improving computation’s societal impact?
Addressing these questions takes time and requires community-
wide HCI eorts and debates. The opportunities outlined initiate
such eorts and debates.
6 OPPORTUNITIES FOR
INDIVIDUAL HCI PROJECTS
Seeing policy as integral to HCI’s intellectual endeavors—be it
revealing the impact of existing technologies (“empirical work”), im-
proving this impact by creating new technologies (“design work”), or
speculating the societal impact of emerging technologies (“futuring
work”)—can enhance these endeavors.
6.1 Integrating Policy into Empirical Work
HCI-Policy Synergies and Tensions. Seeing policy as integral to
HCI’s pursuit in human understanding helps us see the connections
and tensions between them. Both policy and HCI communities want
to understand people’s values, behaviors, societal processes, and
interactions with and experiences of emergent technologies. HCI
and policy share empirical methods such as participatory work-
shops [
91
,
97
]. In many cases, they are already collaborating, e.g., in
understanding misinformation on social media [
39
] and promoting
gig worker welfare [61, 76].
CHI ’24, May 11–16, 2024, Honolulu, HI, USA Yang et al.
Nevertheless, tensions may arise between HCI and policy actors’
empirical work, because of their dierent scopes. HCI’s empirical
ndings illustrate situated interactions among a specic combi-
nation of systems, people, policies, and contexts. They mean to
inform system designs that meet the needs of specic stakehold-
ers in particular contexts. In contrast, policy design inuences a
broader set of technologies, people, and interactions, thus requir-
ing larger-scale empirical evidence. This dierence in scope has
led to some frustrations in the HCI community that policymakers
seek dicult or impossible empirical evidence, such as large-scale
evidence of an emergent technology’s societal harm [77].
Low-Hanging Fruits in Addressing HCI-Policy Tensions. By ex-
plicating the tensions between HCI and policy’s human understand-
ing work, we can identify rst steps in alleviating these tensions. For
example, HCI empirical work can communicate the generalizability
of its policy implications more explicitly. Is an observed technolog-
ical harm specic to this particular technology or population? Is it
the concern of global, national, local, or sectoral regulations? Does
it require an update to the spirit, the text, or the implementation of
the law? Answers to these questions are within the reach of existing
HCI empirical methods, yet can address policymakers’ needs more
directly.
HCI can also expand its empirical methods, observing how a
specic policy helps shape various human-computer interactions,
thereby increasing the likelihood of deriving generalizable policy
implications. What limitations of current AI regulations does gener-
ative AI expose? Is a new privacy policy in conict with the social
norms of particular populations, such that they use technologies to
circumvent it?
Expanding HCI’s use of metaphors to communicate policy impli-
cations represents another tangible opportunity. HCI communities
have long used metaphors when communicating a technology’s
aordances and human impact. Less discussed is that metaphors
often emphasize dierent social values and carry rich policy impli-
cations (e.g., Are generative AI models more like Internet search
engines (therefore their data contributors enjoy similar rights) or
black boxes? Is attaching a GPS surveillance device to a car more
like following a car on public roads (therefore does not require a
court-ordered warrant) or more like trespassing on one’s private
property? [
2
]). In tech policy discussions, various interest groups
often debate the choice of metaphors, because this choice anchors
how policymakers understand a technology’s aordances and de-
cide whether it needs new policy and legal frameworks [
93
]. There
is a ready opportunity for HCI to improve its communication with
policymakers via metaphors. It can help link HCI’s contextual em-
pirical insights to broader policy contexts, using a language that
they are familiar with.
HCI’s empirical work might also take advantage of the fact that
law and policy dier across geographic scales and jurisdictions,
linking local observations with broader policy implications. Take,
for instance, the data breach notication laws in the U.S. Rather
than a single national law governing data breaches, this set of patch-
work rules was created and passed on a state-by-state basis over
sixteen years (2002-2018) [
6
]. This oers a distinctive opportunity
for HCI empirical work. On one hand, HCI can conduct natural ex-
periments, comparatively analyzing how technologies and dierent
laws and their implementations play out dierently. On the other
hand, because companies and stakeholders have the incentive to
comply with the strictest local policies that impact their technolo-
gies, HCI can leverage policy elsewhere to promote change in their
locale of interest.
Emergent HCI-Policy Trans-Disciplinary Methods. Seeing
policy as integral to HCI also allows us to imagine not-yet-existent
empirical methods to study system-people-policy interaction.
For example, we see an opportunity for empirical research on
successful technology and policy designs. Good designs of tech-
nologies, social practices, and policies fade into the background
of everyday life, making them more dicult to observe (e.g., via
stakeholder interviews). Presently, these success stories can oer
valuable lessons, yet are almost entirely absent from HCI literature.
A key aim here is to make what is generally invisible to lay people
visible, to allow technology designers, policymakers, users, and
other stakeholders to reect on the kinds of system-people-policy
interaction design they would nd meaningful and actionable.
Next, if HCI is ready to embrace passive, observational empirical
work on system-people-policy interaction, are we ready to accept
research that creates probes to study tech-policy interactions?
Research through Litigation.
Kirkham
[53]
is a computer scientist with seven years of
experience bringing legal cases to Courts and Tribunals.
He does so for the purpose of (1) understanding how the
law and legal system operates on the ground and, secon-
darily, when possible, (2) changing the law. He named this
approach “Research through Litigation.”
The idea underlying this approach is that the law is a set
of ill-dened rules that variably apply to changing real-
world circumstances. We cannot understand how a law
works simply by reading the law on paper. Further, good
laws often fail because of a lack of institutional capacity to
implement them.
To understand this gap between law on paper and law on
the ground, Kirkham carefully picked legal cases that are
good “test cases.” The legal process allows him to argue
for his case, observe people’s reactions, and, in the pro-
cess, generate voluminous documents (e.g., detailed cor-
respondence within tech companies, expert testimonies)
unlikely to be obtained through other means. He then an-
alyzes this process using autoethnography and document
analysis methods. This approach has surfaced “some seri-
ous (and somewhat surreal) concerns with the operation of
the justice system” and indeed changed the law on several
occasions [53].
People’s experience of technology is highly context-dependent.
HCI empirical researchers routinely use artifacts (e.g., technology
probes, cultural probes) to give users and stakeholders a taste of
the future. In so doing, researchers gain better insights into how
people might interact with future technologies and derive more
informative design implications.
People’s experience of policies is also highly context-dependent.
The proposal of Research through Litigation highlights this com-
plexity that most prior HCI-policy research neglects. In a sense, the
legal cases Kirkham created are analogous to technology probes.
The Future of HCI-Policy Collaboration CHI ’24, May 11–16, 2024, Honolulu, HI, USA
They mean to reveal how law works on the ground, and on occa-
sion, even improve the law. In this light, does this seemingly radical
method merit more consideration? Can HCI eectively simulate
how courts operate, as a way to assess the eects of new tech and
new policy on the ground?
6.2 Integrating Policy into Design Work
HCI-Policy Synergies and Tensions. As mentioned earlier (§2.2),
HCI and policy design share many synergies: Both seek to improve
technologies’ human impact. Both must thoughtfully shape their
design processes, because design processes anchor design outcomes.
Both must navigate the tensions between design goals and technical
feasibility constraints [
8
,
40
]. HCI and policy designers already
share formal methods such as participatory design [25, 91].
Seeing policy as integral to HCI’s design inquiries illuminates
additional connections between them. For example, we realize that
both HCI and policy use service design methods [
48
]. Although
service designers do not always mention the word policy, they
routinely design computational systems, organizational workows,
and related corporate or platform policies in tandem [
33
,
82
]. When
service designers specify a freemium model for a new app, they
specify policies around who can access what quality of service
at what price, a simple micro-economic policy. They shape user
experience by innovating system-human-policy interactions.
Other forms of HCI-policy design collaboration are also emerg-
ing. For example:
•
Designing public service/policy. The boundary between designing
computational systems for government agencies and shaping
public policy is porous. Therefore, many HCI design projects in
the public sector involve policy design. For example, researchers
redesigned the U.S. Postal Service (USPS) manual and, in doing
so, reshaped policies around how small businesses can access
postal services, alleviating their legal burdens [49].
•
Designing both computational systems and policies based on a
shared strong concept. For example, FAccT publications on what
“fairness” [
13
] or “representativeness” [
21
] means catalyzed syn-
ergistic designs of AI systems and their regulations.
•
Designing policy evaluation metrics. For example, Jin
[46]
collabo-
rated with legal professionals to dene evaluation criteria for an
algorithm in the U.S. criminal legal system. This design bridged
HCI and policy experts’ notions of a good algorithm and a good
legal system.
•
Designing legally mandated tech design processes. Some HCI re-
searchers promoted HCI’s human-centered design processes to
become legally required or recommended technology design pro-
cesses [24].
Tensions can also arise between technology and policy design,
for example, due to documentation dierences. HCI and policy
communities dier signicantly in how they document, disseminate,
and attribute their designs. One communicates design ideas via
computational artifacts, demos, and pictorials. The other uses policy
white papers and law reviews. As a result, social connections (e.g.,
an HCI designer taking a gap year to work at government agencies)
can seem like the only route to simultaneously engage in technology
and policy design.
Secondly, coordinating HCI and policy design processes is chal-
lenging, because of their multiple and disparate temporal patterns.
Innovations in HCI system design can occur rapidly, but may also
progress slowly, for example, when developing large computational
infrastructures. Policy change can culminate over decades, but can
also suddenly accelerate, for example, when public interests surge
(e.g., on regulating generative AI [
41
]) or certain public events occur
(e.g., the outbreak of COVID-19 [4]).
Lastly but crucially, power dierences complicate synergistic
HCI-policy designs. Corporate and public policies can shape compu-
tational system designs and user interactions with their authority
and power. The reverse is less true.
Low-Hanging Fruits in Addressing HCI-Policy Tensions. By
explicating the tensions between HCI and policy design, we can
identify practical solutions to address them. For example, improving
documentation of HCI-policy design collaborations. Prominent HCI
research venues can contribute to this by accepting coordinative
tech and policy designs into their proceedings.
Other near-future solutions can further support and popular-
ize the emergent forms of HCI-policy design collaboration listed
above. For example, we see an opportunity to extend the practice
of designing legally mandated tech design processes into the con-
text of privacy by design. While policy documents broadly call for
"privacy by design", it is not always clear how that should be done.
Consequently, legal assessment and risk management practices
ended up operationalizing these design tasks [
94
]. HCI researchers
have already created many design methods to address privacy, and
can more explicitly connect their work with the goals of privacy
by design as articulated in policy documents. More ambitiously,
HCI researchers can actively promote HCI design methods (e.g.,
user-centered design, value-sensitive design) as equally essential
approaches, on par with legal and risk management strategies in
privacy by design.
HCI designers can strategically align their tech design goals
with policymaking timelines. Upon technical and/or social change,
HCI’s novel system design contributions can help ll in the tempo-
rary policy vacuums [
65
], bridging the gap between what the law
protects and promotes and what people believe the law ought to
protect or promote [
11
]. In doing so, HCI’s system designs can help
protect people and circumstances that the law fails to protect [8].
Emergent HCI-Policy Trans-Disciplinary Methods. Seeing pol-
icy as integral to HCI sets the stage for HCI and policy designers to
collaboratively design technologies, social practices, and policies
in tandem, creating an artful tech-policy-human interplay. Service
designers already do so to an extent, though typically within the
connes of one company’s technologies and policies [
20
,
29
]. There
is an open opportunity to expand the emergent practice of simulta-
neous tech-and-policy design to larger-scale design contexts.
Sandhaus et al
. [75]
proposed iCAPS (Integrative Proto-
typing of City Environments, Autonomous Vehicle Behav-
iors, and Policies), a simulation-based prototyping platform.
Build upon the digital twin of a city, this platform simu-
lates and visualizes how design changes to autonomous
vehicles’ (AVs’) driving behaviors, city design, and related
policies might inuence city design goals (e.g., road safety,
CHI ’24, May 11–16, 2024, Honolulu, HI, USA Yang et al.
neighborhood equity, pollution). Via simultaneous AV-city-
policy design, this platform promises to catalyze safer, more
equitable, and more sustainable future cities [35].
This platform is also a boundary object. AV, city, and policy
design decisions belong to many dierent private corpo-
rations and government agencies. Although proven ben-
ecial [
72
], coordinating these design decisions remains
challenging in practice. The iCAPS platform addresses
this challenge by bringing AV, city, and policy designers
together and moderating their design actions.
This emergent method embodies the idea of designing technol-
ogy and policy simultaneously, and proposes one concrete way of
operationalizing it. It raises at least two sets of useful questions.
This method highlights the power struggles among designers
when they design technology and its many related policies in tan-
dem. Systemic thinking is integral to HCI design expertise. It al-
lows designers to grasp how various laws, policies, technologies,
and contextual factors can collectively inuence people’s experi-
ences (Figure 1). However, it would be naive to think that tech
designers can re-design related laws and policies without multiple
policy experts at the table. Must designers convene all relevant
policymakers to the same table, in order to design tech and the
many policies it involves in tandem? If that is not feasible, which
laws and policies should HCI designers prioritize?
This method also underscores the challenges of evaluating system-
people-policy interaction design. While HCI design typically iter-
ates and learns from failures, a failed public policy can have irre-
versible consequences [
73
]. The proposed prototyping platform
addresses this challenge through simulation, computationally pre-
dicting societal outcomes for each AV-city-policy design. This con-
trasts with HCI’s traditional approach of evaluating designs with
real stakeholders. Do HCI communities support this shift?
6.3 Integrating Policy into HCI Futuring Work
HCI-Policy Synergies and Tensions. Both policy and HCI com-
munities want to foresee how emergent technological capabilities
might interact with people and societies, cause disruptions, and
create needs for new computational system designs, social prac-
tices, and/or policies. To do so, the eld of HCI has methods such
as speculative design and consequence scanning [
3
,
50
]. Policy ac-
tors have methods such as forecasting [
76
] and the Precautionary
Principle [55].
Yet at least two dierences b etween HCI’s and p olicy’s approaches
can hinder their collaboration. Prior research often criticized HCI’s
policy engagements being too reactive [
43
]. These hindrances might
explain why.
First is their dierent views of risks. HCI’s speculative designs
indicate qualitative risks, often without indicating the level of ur-
gency. Their goal is to provoke discussion and providing cautionary
tales for new technologies’ adoption and (mis)use. Risk and urgency
levels, however, are essential for policy actors’ futuring work: Is a
new technology’s societal harm catastrophic but unlikely (like nu-
clear war), slowly unfolding but inevitable (like climate change), or
so catastrophic and inevitable that policy should forbid it outside of
research labs if not entirely (like genetic editing of deadly diseases)?
The answer to this question is crucial for policy-making [
45
], yet
fall outside of existing HCI futuring methods.
Second is their dierent views of design. As we have hinted
at earlier in the paper, while speculative design is a process of
open-ended knowledge inquiry [
94
], some legal scholarship sees
design as merely the implementation of a given policy requirement.
Many policymakers engaged designers only after they had identi-
ed a policy goal (e.g., ensure citizens’ right to privacy) and had
translated it into system requirements (e.g., providing individual
control over personal information by increasing consumer notice
and choice [
34
]). They engaged HCI designers only to implement
Civil Rights Laws
E.g., data ownership and privacy, accessible ground tra nsportation laws in th e ADA.
AV Motion Design
Algorithms that instruct how autonomous
vehicles move
Trac Regulations
Laws and policies that regulate how vehicles
and people move
Transportation Regulations
Laws and policies that set rules and incentives
related to road use and vehicle use
AV Service Design
Service designs that set rules and incentives
related to autonomous vehicles use
Urban/Rural Infrastructure Design
City planning, sensors and “smart city” design, road and highway plans, design of parking space, la nes for specific
vehicles, signs etc.
Restrict
what AV
designs
are possible
Create needs & data
evidence for new
urban/rural design
Restrict
what AV
designs
are possible
Create needs for
new regulations
Incentivize or
mandate certain AV
ownerships/uses/
services while
restricting oth ers
Create needs for
new regulations
Policies
incentivize or
mandate certain
urban designs
while restricting
others
Policies
enforce good
urban design
choices
Restrict
what service
designs
are possible
Create opportunities and
trials for novel public
transportation service
Figure 1: The many ways autonomous vehicle (AV) behavior design, city design, and law and policy
interact in the U.S. legal context. They enable (green arrows) and constrain (red arrows) each other [
75
].
The Future of HCI-Policy Collaboration CHI ’24, May 11–16, 2024, Honolulu, HI, USA
these requirements (e.g., design the standardized cookie opt-out
menus [84]).
Low-Hanging Fruits in Addressing HCI-Policy Tensions. By
explicating the connections and tensions between HCI and policy’s
futuring methods, we start to identify easy ways to strengthen the
connections and alleviate tensions.
For example, HCI’s speculative design work can articulate its
<Implications for Policy> more explicitly and concretely. What
emergent technologies (e.g., facial recognition AI) might require
new methods to meet old policy goals? In what ways may cur-
rent laws and policies fall short in addressing a problem? What
values that people hold do not t into existing legal frameworks
(e.g., changing privacy norms)? How might policymakers dene a
new policy problem space (e.g., “dark patterns” [
37
]), based on the
emergent socio-technical issues this work has identied? Are new
policy initiatives necessary [62]?
Explicating policy implications of HCI’s speculative design work
can potentially increase its policy impact. In addition, this approach
can showcase the value of speculative design throughout the policy-
making process, gradually correcting misconceptions about HCI’s
design expertise among some policy actors. Imagine, if HCI re-
searchers had dened the new social issues and policy needs that
chatGPT would entail, within the four years after its neural ar-
chitecture was rst published in academia and before its public
release. These insights might just tackle the challenge of “bringing
policymakers to the table” [77].
Emergent HCI-Policy Trans-Disciplinary Methods. Seeing
policy as integral to HCI’s futuring work also gives us license
to imagine entirely new types of futuring research. For example,
speculative design practice might embrace speculations around
law and policy. It can speculate about the social and technical
implications of an emergent policy proposal. It can also envision
new uses and forms for emerging technologies, new categories of
law and policies, and new social practices. After all, these factors
and the interactions among them simultaneously shape emergent
technologies’ societal impact.
More ambitiously, we see opportunities in exploring new tech-
nological risk forecasting methods that combine HCI’s qualitative,
provocation-oriented approaches with policy’s semi-quantitative,
action-oriented approaches. To help provoke our imaginations
around these opportunities, we describe one such method emer-
gent from our workshop; a method that some may consider too
progressive, or too tilted towards policy methods and not enough
towards HCI. Presenting this method as a provocation, we ask:
Suppose we (HCI communities) are ready to embrace policy as an
integral part of its futuring work. Are we also ready to embrace
political science’s view of rigor in forecasting?
AI Capability Forecasting:
Dardaman and Gupta
[23]
proposed to quantitatively fore-
cast AI capability growth, in order to change the reactionary
stance HCI researchers, policymakers, and organizations of-
ten found themselves in when regulating AI. Such forecasts,
and the process of generating and debating them, can help
them build mid- and long-term strategies for addressing
AI’s societal impact. The quantitative forecasting method
has three steps:
1.
Dene a technology capability prediction problem with
clear evaluation criteria;
2.
Invite many researchers and forecasters with a variety
of relevant expertise to submit their quantitative pre-
dictions, along with documentations of their methods,
assumptions, and uncertainty measures;
3.
Aggregate the forecasts by examining their degree of
consensus. As a secondary step, some researchers bring
forecasters together, enabling them to challenge each
others’ assumptions, sharpen their methods, and collab-
oratively converge on the most likely predictions.
This method challenges HCI’s assumption that speculations of
technology harms are only qualitative. To inform policy actions,
analysis of how likely or how soon the harm will become a reality
is critical. In order to impact policy, is HCI’s speculative design
work open to alternative criteria of rigor in its speculations?
Throughout Chapter 6, we illustrated that seeing policy as in-
tegral to HCI’s intellectual endeavors can enhance these endeav-
ors. Examining existing HCI literature (including ones previously
considered at the fringes of HCI) through this new lens, we can
begin to see inherent tensions between HCI and policy methods,
identify pragmatic approaches to alleviate these tensions, and delib-
erate on emergent methods that address these tensions in thought-
provoking ways. Going forward, we encourage HCI communities
to together foster a mosaic of distinctively HCI methods and knowl-
edge contributions that blend system, human, and policy expertise
to various degrees and in diverse ways, creating robust, grass-root
connections between the two elds. Chapter 6 outlines an initial
draft of this new landscape.
7 OPPORTUNITIES IN ENHANCING HCI’S
COLLECTIVE IMPACT ON POLICY
Now, we turn our attention to HCI-policy collaboration at a com-
munity level. We previously proposed that, by seeing policy as
integral to their intellectual concerns, HCI communities will coor-
dinate their diverse policy engagements, thereby amplifying HCI’s
collective impact on policy outcomes. In this chapter, we outline
four emergent opportunities for such coordination by synthesising
existing HCI and policy literature.
7.1 When (Not) to Policy
We envision a future where HCI communities strategically choose
when and when not to use policy to address computation’s societal
issues. A crucial step toward realizing this vision is establishing a
principled understanding of policy’s aordances and limitations,
compared with other HCI’s tools. There is an under-explored area
in HCI research and a clear opportunity for future work.
On the one hand, HCI can further exploit the aordances of pol-
icy. For example, current HCI-policy work focused overwhelmingly
on regulating technology harms. Less discussed is how policy can
also enable new actions. For example, copyright law attempts to
balance protecting creators from intellectual property harms, while
also encouraging new innovations for social good through fair use.
How might HCI leverage policy to promote the positive impact of
technologies?
CHI ’24, May 11–16, 2024, Honolulu, HI, USA Yang et al.
On the other hand, policy interventions are not a cure-all, yet
their limitations are overlooked in current HCI literature. They can
be prone to their own forms of inequities and bias [
9
]. Moreover,
law and policy solutions promote incremental changes within an
existing legal and institutional structure. They typically do not
“change the system,” particularly not overnight [
17
]. Depending on
the degree and kind of social change HCI hopes to create, some-
times, other approaches (e.g., community organizing) might be
more eective than policy interventions.
Explicating the limitations of policy can also reveal new oppor-
tunities for technology design. For example, HCI researchers often
criticize that capitalist goals drive many technology product de-
signs and cause ethical harm. They then seek remedies from policy
interventions. However, law and policy can be just as enwrapped
in promoting or serving institutions of capitalism [
22
]. In order
to improve technologies’ societal impact, HCI sometimes needs
to consider designing technologies that help the disenfranchised
populations ght unjust policies [10].
7.2
Which Mode of Policy Engagement & When
We envision a future where HCI communities will wisely choose
the extents and modes of policy engagement, based on the specic
technology and social issue at hand. A crucial step toward realizing
this vision is to develop a principled understanding of how the most
productive approach to policy engagement diers according to
context. For example, when is appending a thoughtful <Implications
for Policy> to a classic HCI empirical study already sucient?
In what contexts is establishing deep, ongoing partnerships with
policymakers not just benecial, but necessary?
The history of HCI—more specically, the three waves of HCI [
15
]—
can serve as an initial scaolding for this discussion. This history
seems to suggest that as a computing system becomes more deeply
and broadly integrated into societal processes, addressing its impact
requires deeper collaboration between HCI and policy.
•
The rst wave of HCI focuses on making computers more e-
cient and functional for computer experts. These eorts involved
minimal policy considerations.
•
The second wave of HCI focuses on making personal comput-
ers easier to use and more accessible to everyday users. These
eorts catalyzed the ISO usability standards and accessibility
laws, which in turn enhanced later HCI work. Here, HCI and
policy work operated asynchronously. Appending a thoughtful
<Implications for Policy> to an HCI empirical study on usability
or accessibility could go a long way.
•
The third wave of HCI focused on weaving computing into daily
life through smartphones, tablets, and wearable devices. Around
this time, HCI privacy scholars like Lorrie Cranor drafted national
privacy policies at the FTC and designed standardized cookie opt-
out menus [
24
]. State laws later recommended these designs [
84
].
Here, while HCI and policy work remained somewhat asynchro-
nous, technology and policy design processes started to overlap.
Do the three waves of HCI eectively stratify technologies’ needs
for HCI-policy collaborative work? Is there a need for additional
or more ne-grained categories? These questions provide fertile
ground for future research.
For example, the societal impact of interactive cyber-infrastructure
appears to demand particularly close HCI-policy partnerships [
30
],
and therefore might merit its own category. Interactive cyber-
infrastructure refers to interactive systems that not only host, but
govern various interactions on a societal-level scale
2
. For instance,
social media platforms whose algorithms govern online discourse,
gig work platforms whose algorithms govern work assignments,
pervasive sensing networks that govern how smart cities operate,
AI Foundation Models (FMs) such as chatGPT that govern down-
stream AI innovations, and virtual reality (VR) “universes” where
governments oer public services [
7
]. Understanding and improv-
ing the societal impact of cyber-infrastructure necessitate close
HCI-policy partnerships, for three reasons.
•
Because of the immense scale of these systems, HCI’s conven-
tional empirical methods may struggle to trace or prove their
societal impact. Law and policy methods (e.g., Research through
Litigation) might help;
•
Because of the immense power the cyber-infrastructure owners
wield (think the power of OpenAI), HCI designers heavily rely
on policies that mandate system transparency and/or access, in
order to do human-centered design work. Designing one more
socially benecial GPT prompt is simply not as impactful as
devising a way of reducing the creation of harmful prompts. The
latter necessitates policy interventions.
•
Because these systems are infrastructures, they shape how HCI
researchers and practitioners work. Should CHI keep publishing
novel GPT applications, even though they might be soon regulated
out of existence? Questions like this highlight the necessity of
HCI-policy considerations from the outset.
7.3 How to Maximize HCI’s Collective Impact
We envision a future where HCI can assemble the dierent sub-
sets of its work to strategically inuence dierent stages of policy-
making, and to inform dierent policy actors [
48
,
52
]. When sud-
den momentum for policy change appears, HCI communities can
quickly assemble their diverse relevant work to seize the opportu-
nity.
Preparing the ground for policy change. Even policy changes
that seem to occur “suddenly” have longer histories of community
discussion and debate. This period is an opportune time for cultivat-
ing relationships with policy actors, engaging in policy discussions
that occur in the background of everyday life, and incubating HCI
empirical and design work.
HCI researchers and practitioners can become more actively
engaged in everyday policy discussions, for example, by respond-
ing to regulatory agencies’ requests for public comment [
1
] and
writing Op-Ed articles (and other forms of tech journalism). In
these discussions, HCI professionals can share metaphors that help
policymakers understand the aordance of emergent technologies,
oer empirical evidence of technological harm as expert evidence
for future policies, recommend human-centered design methods to
be legally mandated or recommended, and propose value-sensitive
2
To be clear, here we use the term "interactive cyber-infrastructure" to denote a distinct
category of computational systems, rather than as an analytical lens on the sociotech-
nical systems that shape society. We use the word “infrastructure” dierently than in
Science and Technology Studies (STS) and cultural studies (e.g., [30, 31, 79]).
The Future of HCI-Policy Collaboration CHI ’24, May 11–16, 2024, Honolulu, HI, USA
metrics for evaluating future policies. All this work can prepare the
ground for future policy change.
In addition, the potentially long “incubation” period of policy
change is also opportune for long-term HCI research, such as con-
ducting participatory design workshops and comparatively ana-
lyzing how dierent tech-and-policy designs play out in dierent
locales.
Catalyzing a groundswell of public interest and policy de-
mand. Law and policy can change quickly in response to cur-
rent events (e.g., COVID-19 [
4
]), technical developments (e.g., chat-
GPT [
90
]), and other changes. Yet that does not mean HCI has to
wait passively for a pandemic to see windows of opportunity for
policy change open up.
HCI’s speculative, provocative, and participatory design work
can catalyze a groundswell of public opinion around HCI issues,
creating a new policy demand. Critical HCI is already somewhat
successful in this regard [
91
]. Additional opportunities exist for
HCI to engage interests groups, corporations, and policymakers
more pointedly. The aim is to align the stars: When “a problem is
recognized, a solution is available, and the political climate happens
to be right”, a window of opportunity for substantial policy change
opens [52].
Seizing sudden momentum of policy change. When a window
of opportunity for policy change opens up, the heightened levels
of attention are eeting [
52
]. HCI communities should be prepared
to seize the political momentum.
HCI researchers can prepare rapid policy responses by re-using
their existing work, but adding a “hook” that connects it to the
current events and political momentum. <Implications for Policy>
sections of past HCI futuring work, literature reviews summarizing
existing empirical evidence, and policy proposals tested in prior
participatory workshops are all valuable, reusable resources here.
These responses can take the form of op-eds, memos, policy white
papers, public agency comments, and more.
7.4 How Might HCI Institutions Help
Finally and importantly, we call for HCI institutions to nurture the
ties between HCI and policy communities and to amplify HCI’s
collective voice in the policy world.
Including policy education in HCI; including HCI in policy
education. We envision a future where policy-making and policy
implementation become a standard element of HCI education. This
education can cultivate future HCI researchers and practitioners
who can navigate layers of government agencies, navigate laws
of intricately overlapping jurisdictions, connect with policymak-
ers, and integrate policy expertise into their own work. We also
see an opportunity for HCI to become part of legal and political
science education. Today, many leading law schools already oer
curriculum on AI ethics, as do many HCI programs. Harnessing and
expanding these existing connections incubate future HCI-policy
collaborations of all sorts.
Celebrating small wins. From accumulating research evidence,
to building relationships with policymakers, to acquiring policy
expertise, HCI researchers and practitioners’ journey to policy im-
pact is arduous. Moreover, such impacts are often invisible: Citing
academic references or attributing individual researchers is not
a convention in legal or public policy documents. How can HCI
institutions (such as CHI and CSCW) acknowledge and celebrate
small wins in HCI researchers’ policy engagement?
Consider: Can CHI proceedings accept insightful literature re-
views that synthesize prior HCI works’ <Implications for Policy>
into actionable policy proposals? Can CHI oer a short paper track
where researchers share their responses to regulatory agencies’
calls for comments? When public agencies or policymakers call
for scientic evidence, they often require a response within days
or weeks [
24
]. In parallel to peer-reviewed publications, can HCI
conferences create channels for faster-paced policy discourses and
contributions? Such small recognitions and incentives have the
potential to signicantly boost policy engagements, particularly
for early-career professionals in HCI.
Amplifying HCI’s collective voice in the policy realm. HCI
institutions can help strengthen the collective ties between HCI
and policy communities. For example, the registration costs for HCI
conferences may be dicult for public servants to justify; are there
ways to make the HCI community more accessible to these policy
audiences? Can CHI sets up booths for policymakers and public
servants (just like it does for recruiters), helping HCI researchers
connect with them?
HCI institutions can help amplify HCI communities’ collective
voices in the policy realm. For example, HCI can oer social infras-
tructures that enable HCI professionals to collaborate on public
comments, statements, and policy recommendations. This approach
can help more evenly distribute the policy engagement work across
community members, and can amplify the collective voice of HCI
communities.
8 CLOSING REMARKS
The eld of HCI has a history of self-improvement: identifying a
problem, creating a new way of working, and then using evidence
of success to argue for a change in the community. Early computer
systems caused chaos when they rst entered workplaces, facing
users who were not computer scientists. Early HCI researchers
(before even calling themselves HCI) created a new way of design-
ing technologies in response, that is, user-centered design. When
computing moved out of the workplace and into people’s everyday
lives, users became consumers, and they had choices. In response,
HCI practitioners worked to improve the situated experience of
technologies, and researchers dug in and unpacked the very concept
of “experience”.
All this bodes well for the changes facing our community today.
In current technology discourses, people are not only users and
consumers. They are members of communities, citizens, and par-
ticipants in lived democracy. Technologies are not only systems.
They are socio-technical platforms and cyber-infrastructure that
govern societal processes. People’s “experience” of technologies
goes beyond eciency or pleasure, but concerns the future of work,
equity, sustainability, and more. Within such a milieu, we argue for
a change in our community. We argue that HCI should re-examine
the relationship among systems, people, and policy, and reect on
how HCI wants to position itself in this relationship. This paper
provides an additional perspective to this unfolding dialogue.
CHI ’24, May 11–16, 2024, Honolulu, HI, USA Yang et al.
ACKNOWLEDGMENTS
Qian Yang’s eort to help organize and participate in the work-
shop was partially supported by National Science Foundation under
grants IIS-2212431, IIS-2313077, and Schmidt Futures’ AI2050 Early
Career Fellowship. Richmond Y. Wong’s eort was supported by a
Georgia Tech Ivan Allen College Small Grant for Research (SGR).
Sabine Junginger’s work on integrating policy and service design
is supported by the Swiss National Science Foundation SINERGIA
CRSII5 189955 VA-PEPR (Voice Assistants, People Experiences, Prac-
tices and Routines). John Zimmerman’s eort to help organize and
participate in the workshop was partially supported by National
Science Foundation under grants IIS-2007501 and IIS-2112633. Any
opinions, ndings, conclusions, or recommendations expressed in
this material are those of the author(s) and do not necessarily reect
the views of the National Science Foundation.
REFERENCES
[1]
2011. https://www.federalregister.gov/uploads/2011/01/the_rulemaking_process.
pdf
[2]
2012. United States v. Jones, 565 U.S. 400 (2012). https://supreme.justia.com/
cases/federal/us/565/400/
[3] 2017. https://www.gov.uk/guidance/open-policy-making-toolkit
[4] 2020. https://crsreports.congress.gov/product/pdf/LSB/LSB10511
[5]
2021. https://www.ftc.gov/system/les/documents/public_events/1586943/dark_
patterns-workshop- bios.pdf
[6]
2022. https://www.ncsl.org/technology-and-communication/security-breach-
notication-laws
[7]
Accenture. 2023. Government enters the metaverse. https://www.accenture.
com/us-en/insightsnew/us- federal-government/technology- vision- 2022
[8]
Mark S Ackerman. 2000. The intellectual challenge of CSCW: the gap between
social requirements and technical feasibility. Human–Computer Interaction 15,
2-3 (2000), 179–203.
[9]
Kendra Albert and James Grimmelmann. 2023. Do the Right Thing. Commun.
ACM 66, 5 (2023), 18–20.
[10]
Tran Anh. 2023. Adversarial Engagements Between Technology and Policy. In
Workshop on Designing Technology and Policy Simultaneously: Towards A Research
Agenda and New Practice.
[11]
Kenneth A Bamberger and Deirdre K Mulligan. 2011. Privacy on the Books and
on the Ground. Stanford Law Review (2011), 247–315.
[12]
Justin Bariso. 2016. Why does uber keep breaking the law? because they’re
disrupting, of course. https://www.inc.com/justin-bariso/why- does-uber-keep-
breaking-the- law-because-theyre- disrupting-of- course.html
[13]
Solon Barocas, Moritz Hardt, and Arvind Narayanan. 2017. Fairness in machine
learning. Nips tutorial 1 (2017), 2017.
[14]
Eli Blevis. 2007. Sustainable interaction design: invention & disposal, renewal
& reuse. In Proceedings of the SIGCHI conference on Human factors in computing
systems. 503–512.
[15]
Susanne Bødker. 2015. Third-wave HCI, 10 years later—participation and sharing.
interactions 22, 5 (2015), 24–31.
[16]
Rishi Bommasani, Drew A. Hudson, Ehsan Adeli, Russ Altman, Simran Arora,
Sydney von Arx, Michael S. Bernstein, Jeannette Bohg, Antoine Bosselut, Emma
Brunskill, Erik Brynjolfsson, Shyamal Buch, Dallas Card, Rodrigo Castellon,
Niladri Chatterji, Annie Chen, Kathleen Creel, Jared Quincy Davis, Dora Dem-
szky, Chris Donahue, Moussa Doumbouya, Esin Durmus, Stefano Ermon, John
Etchemendy, Kawin Ethayarajh, Li Fei-Fei, Chelsea Finn, Trevor Gale, Lauren
Gillespie, Karan Goel, Noah Goodman, Shelby Grossman, Neel Guha, Tatsunori
Hashimoto, Peter Henderson, John Hewitt, Daniel E. Ho, Jenny Hong, Kyle Hsu,
Jing Huang, Thomas Icard, Saahil Jain, Dan Jurafsky, Pratyusha Kalluri, Siddharth
Karamcheti, Geo Keeling, Fereshte Khani, Omar Khattab, Pang Wei Kohd, Mark
Krass, Ranjay Krishna, Rohith Kuditipudi, Ananya Kumar, Faisal Ladhak, Mina
Lee, Tony Lee, Jure Leskovec, Isabelle Levent, Xiang Lisa Li, Xuechen Li, Tengyu
Ma, Ali Malik, Christopher D. Manning, Suvir Mirchandani, Eric Mitchell, Zanele
Munyikwa, Suraj Nair, Avanika Narayan, Deepak Narayanan, Ben Newman,
Allen Nie, Juan Carlos Niebles, Hamed Nilforoshan, Julian Nyarko, Giray Ogut,
Laurel Orr, Isabel Papadimitriou, Joon Sung Park, Chris Piech, Eva Portelance,
Christopher Potts, Aditi Raghunathan, Rob Reich, Hongyu Ren, Frieda Rong,
Yusuf Roohani, Camilo Ruiz, Jack Ryan, Christopher Ré, Dorsa Sadigh, Shiori
Sagawa, Keshav Santhanam, Andy Shih, Krishnan Srinivasan, Alex Tamkin, Ro-
han Taori, Armin W. Thomas, Florian Tramèr, Rose E. Wang, William Wang,
Bohan Wu, Jiajun Wu, Yuhuai Wu, Sang Michael Xie, Michihiro Yasunaga, Ji-
axuan You, Matei Zaharia, Michael Zhang, Tianyi Zhang, Xikun Zhang, Yuhui
Zhang, Lucia Zheng, Kaitlyn Zhou, and Percy Liang. 2021. On the Opportunities
and Risks of Foundation Models. arXiv:2108.07258 [cs.LG]
[17]
Karrigan S Bork. 2019. An evolutionary theory of administrative law. SMU L.
Rev. 72 (2019), 81.
[18]
Jed R. Brubaker and Vanessa Callison-Burch. 2016. Legacy Contact: Designing and
Implementing Post-Mortem Stewardship at Facebook. In Proceedings of the 2016
CHI Conference on Human Factors in Computing Systems (San Jose, California,
USA) (CHI ’16). Association for Computing Machinery, New York, NY, USA,
2908–2919. https://doi.org/10.1145/2858036.2858254
[19]
Matt Burgess. 2023. Criminals Have Created Their Own ChatGPT Clones. https:
//www.wired.com/story/chatgpt-scams- fraudgpt-wormgpt- crime/
[20]
Alissa Centivany. 2016. Policy as Embedded Generativity: A Case Study of
the Emergence and Evolution of HathiTrust. In Proceedings of the 19th ACM
Conference on Computer-Supported Cooperative Work & Social Computing (San
Francisco, California, USA) (CSCW ’16). Association for Computing Machinery,
New York, NY, USA, 926–940. https://doi.org/10.1145/2818048.2820069
[21]
Kyla Chasalow and Karen Levy. 2021. Representativeness in statistics, politics,
and machine learning. In Proceedings of the 2021 ACM Conference on Fairness,
Accountability, and Transparency. 77–89.
[22]
Ernesto Dal Bó. 2006. Regulatory capture: A review. Oxford review of economic
policy 22, 2 (2006), 203–225.
[23]
Emily Dardaman and Abhishek Gupta. 2023. Asking Better Questions – The Art
and Science of Forecasting: A mechanism for truer answers to high-stakes ques-
tions. In Workshop on Designing Technology and Policy Simultaneously: Towards
A Research Agenda and New Practice. arXiv:2303.18006 [cs.CY]
[24]
Janet Davis, Harry Hochheiser, Juan Pablo Hourcade, Je Johnson, Lisa Nathan,
and Janice Tsai. 2012. Occupy CHI! Engaging U.S. Policymakers. In CHI ’12
Extended Abstracts on Human Factors in Computing Systems (Austin, Texas,
USA) (CHI EA ’12). Association for Computing Machinery, New York, NY, USA,
1139–1142. https://doi.org/10.1145/2212776.2212406 Note: Panel slides can be
found at https://web.archive.org/web/20170829055308/http://www.cs.grinnell.
edu/~davisjan/chi-us- public-policy/chi2012-panel.pdf . Accessed: 2023-07-31.
[25]
Fernando Delgado, Stephen Yang, Michael Madaio, and Qian Yang. 2023. The
Participatory Turn in AI Design: Theoretical Foundations and the Current State
of Practice. In Proceedings of the 3rd ACM Conference on Equity and Access in
Algorithms, Mechanisms, and Optimization. 1–23.
[26]
Carl DiSalvo, Jonathan Lukens, Thomas Lodato, Tom Jenkins, and Tanyoung Kim.
2014. Making public things: how HCI design can express matters of concern. In
Proceedings of the SIGCHI Conference on Human factors in Computing Systems.
2397–2406.
[27]
Lynn Dombrowski, Ellie Harmon, and Sarah Fox. 2016. Social Justice-Oriented
Interaction Design: Outlining Key Design Strategies and Commitments. In Pro-
ceedings of the 2016 ACM Conference on Designing Interactive Systems (Brisbane,
QLD, Australia) (DIS ’16). Association for Computing Machinery, New York, NY,
USA, 656–671. https://doi.org/10.1145/2901790.2901861
[28]
Kees Dorst and Nigel Cross. 2001. Creativity in the design process: co-evolution
of problem–solution. Design studies 22, 5 (2001), 425–437.
[29]
Brianna Dym, Namita Pasupuleti, and Casey Fiesler. 2022. Building a Pillowfort:
Political Tensions in Platform Design and Policy. Proc. ACM Hum.-Comput.
Interact. 6, GROUP, Article 16 (jan 2022), 23 pages. https://doi.org/10.1145/
3492835
[30]
Paul N Edwards, Steven J Jackson, Georey C Bowker, and Cory Philip Knobel.
2007. Understanding infrastructure: Dynamics, tensions, and design. (2007).
[31]
W. Keith Edwards, Mark W. Newman, and Erika Shehan Poole. 2010. The
Infrastructure Problem in HCI. In Proceedings of the SIGCHI Conference on
Human Factors in Computing Systems (Atlanta, Georgia, USA) (CHI ’10). As-
sociation for Computing Machinery, New York, N Y, USA, 423–432. https:
//doi.org/10.1145/1753326.1753390
[32]
Jodi Forlizzi. 2018. Moving beyond user-centered design. interactions 25, 5 (2018),
22–23.
[33]
Jodi Forlizzi. 2018. Moving beyond user-centered design. interactions 25, 5 (2018),
22–23.
[34] Robert Gellman. 2022. Fair Information Practices: A Basic History-Version 2.22.
Available at SSRN (2022).
[35]
Thomas Krendl Gilbert, Aaron J Snoswell, Michael Dennis, Rowan Mcallister,
and Cathy Wu. 2022. Sociotechnical Specication for the Broader Impacts of
Autonomous Vehicles. ICRA’22 workshop Fresh Perspectives on the Future of
Autonomous Driving.
[36]
Colin M Gray. 2022. Expert Report of Colin M. Gray, Ph. D. on Dark Patterns
(Public Redacted Version). STATE OF ARIZONA, ex rel. MARK BRNOVICH, Attor-
ney General, Plainti, v. Google LLC (2022). https://www.azag.gov/sites/default/
les/2022-09/Expert%20Report%20of%20Colin%20M.%20Gray%2C%20Ph.D..pdf
[37]
Colin M Gray, Yubo Kou, Bryan Battles, Joseph Hoggatt, and Austin L Toombs.
2018. The dark (patterns) side of UX design. In Proceedings of the 2018 CHI
conference on human factors in computing systems. 1–14.
[38]
Louise Møller Haase and Linda Nhu Laursen. 2019. Meaning frames: The structure
of problem frames and solution frames. Design Issues 35, 3 (2019), 20–34.
The Future of HCI-Policy Collaboration CHI ’24, May 11–16, 2024, Honolulu, HI, USA
[39]
Oliver L Haimson, Daniel Delmonaco, Peipei Nie, and Andrea Wegner. 2021.
Disproportionate removals and diering content moderation experiences for
conservative, transgender, and black social media users: Marginalization and
moderation gray areas. Proceedings of the ACM on Human-Computer Interaction
5, CSCW2 (2021), 1–35.
[40]
Michael T Hayes. 2002. The limits of policy change: Incrementalism, worldview,
and the rule of law. Georgetown University Press.
[41]
David Hickton. 2023. Chatgpt could transform society and its risks require quick
regulation. https://thehill.com/opinion/technology/3812597-chatgpt- could-
transform-society- and-its-risks- require-quick- regulation/
[42]
Jason Hong. 2023. What Can the FATE Community Learn from the Successes
and Failures in Privacy? https://cacm.acm.org/blogs/blog-cacm/271991-
what-can- the-fate- community- learn-from- the-successes-and- failures-in-
privacy/fulltext. Accessed: 2023-07-29.
[43]
Steven J. Jackson, Tarleton Gillespie, and Sandy Payette. 2014. The Policy Knot:
Re-Integrating Policy, Practice and Design in CSCW Studies of Social Computing.
In Proceedings of the 17th ACM Conference on Computer Supported Cooperative
Work & Social Computing (Baltimore, Maryland, USA) (CSCW ’14). Association
for Computing Machinery, New York, NY, USA, 588–602. https://doi.org/10.
1145/2531602.2531674
[44]
Erich Jantsch. 1972. Inter-and transdisciplinary university: A systems approach
to education and innovation. Higher education 1, 1 (1972), 7–37.
[45]
Sheila Jasano. 2014. Serviceable truths: Science for action in law and policy.
Tex. L. Rev. 93 (2014), 1723.
[46]
Angela Jin. 2023. Opportunities for Coordinating Policy and Technology Design
for Algorithmic Accountability: Insights from the U.S. Criminal Legal System. In
Workshop on Designing Technology and Policy Simultaneously: Towards A Research
Agenda and New Practice.
[47]
Sabine Junginger. 2013. Design and Innovation in the Public Sector: Matters of
Design in Policy-Making and Policy Implementation. Annual Review of Policy
Design 1 (2013). Issue 1. https://ojs.unbc.ca/index.php/design/article/view/542
[48]
Sabine Junginger. 2016. Transforming Public Services by Design: Re-orienting
policies, organizations and services around people. Taylor & Francis.
[49]
Sabine Junginger. 2016. The USPS domestic mail manual transformation project.
In Transforming Public Services by Design: Re-orienting policies, organizations and
services around people. Taylor & Francis.
[50]
Rob Katz. 2021. How to run a consequence Scanning workshop.
https://medium.com/salesforce-ux/how- to-run- a- consequence-scanning-
workshop-4b14792ea987
[51]
Sarah Kessler. 2013. Uber: When innovation outpaces the law. https://www.
fastcompany.com/3001169/uber-when-innovation- outpaces-law
[52]
John W Kingdon and Eric Stano. 1984. Agendas, alternatives, and public policies.
Vol. 45. Little, Brown Boston.
[53]
Reuben Kirkham. 2023. (Legal Design) Research through Litigation. In Workshop
on Designing Technology and Policy Simultaneously: Towards A Research Agenda
and New Practice. arXiv:2303.14336 [cs.HC]
[54]
Julie Thompson Klein. 1990. Interdisciplinarity: History, theory, and practice.
Wayne state university press.
[55]
David Kriebel, Joel Tickner, Paul Epstein, John Lemons, Richard Levins, Edward L
Loechler, Margaret Quinn, Ruthann Rudel, Ted Schettler, and Michael Stoto. 2001.
The precautionary principle in environmental science. Environmental health
perspectives 109, 9 (2001), 871–876.
[56]
Benjamin Laufer, Sameer Jain, A Feder Cooper, Jon Kleinberg, and Hoda Heidari.
2022. Four years of FAccT: A reexive, mixed-methods analysis of research
contributions, shortcomings, and future prospects. In Proceedings of the 2022
ACM Conference on Fairness, Accountability, and Transparency. 401–426.
[57]
Jonathan Lazar. 2010. Interacting with Public Policy. Interactions 17, 1 (jan 2010),
40–43. https://doi.org/10.1145/1649475.1649485
[58]
Jonathan Lazar. 2015. Public Policy and HCI: Making an Impact in the Future.
Interactions 22, 5 (aug 2015), 69–71. https://doi.org/10.1145/2807916
[59]
Jonathan Lazar. 2017. Let’s strengthen the HCI community by taking a gap year!
Interactions 25, 1 (2017), 20–21.
[60]
Jonathan Lazar, Julio Abascal, Janet Davis, Vanessa Evers, Jan Gulliksen, Joaquim
Jorge, Tom McEwan, Fabio Paternò, Hans Persson, Raquel Prates, Hans von
Axelson, Marco Winckler, and Volker Wulf. 2012. HCI Public Policy Activities
in 2012: A 10-Country Discussion. Interactions 19, 3 (may 2012), 78–81. https:
//doi.org/10.1145/2168931.2168947
[61]
Min Kyung Lee, Daniel Kusbit, Evan Metsky, and Laura Dabbish. 2015. Working
with machines: The impact of algorithmic and data-driven management on
human workers. In Proceedings of the 33rd annual ACM conference on human
factors in computing systems. 1603–1612.
[62]
Joseph Galen Lindley, Paul Coulton, Haider Akmal, and Brandin Hanson Knowles.
2017. Anticipating GDPR in Smart Homes Through Fictional Conversational
Objects. (2017).
[63]
Michael A. Madaio, Luke Stark, Jennifer Wortman Vaughan, and Hanna Wallach.
2020. Co-Designing Checklists to Understand Organizational Challenges and
Opportunities around Fairness in AI. In Proceedings of the 2020 CHI Conference on
Human Factors in Computing Systems (Honolulu, HI, USA) (CHI ’20). Association
for Computing Machinery, New York, NY, USA, 1–14. https://doi.org/10.1145/
3313831.3376445
[64]
Lisa Messeri and Janet Vertesi. 2015. The Greatest Missions Never Flown: Antici-
patory Discourse and the" Projectory" in Technological Communities. Technology
and Culture (2015), 54–85.
[65]
James H Moor. 1985. What is computer ethics? Metaphilosophy 16, 4 (1985),
266–275.
[66]
Lisa P. Nathan and Batya Friedman. 2010. Interacting with Policy in a Political
World: Reections from the Voices from the Rwanda Tribunal Project. Interactions
17, 5 (sep 2010), 56–59. https://doi.org/10.1145/1836216.1836231
[67]
Andy Newman. 2019. I found work on an Amazon website. I made 97 cents an
hour. The New York Times 15 (2019).
[68]
Hope Nudzor. 2009. What is “policy”, a problem: solving denition or a process
conceptualisation. Educational futures 2, 1 (2009), 85–96.
[69]
Billy Perrigo. 2023. Exclusive: Openai lobbied E.U. to water down AI regulation.
https://time.com/6288245/openai-eu- lobbying-ai- act/
[70]
B Guy Peters. 2021. Advanced introduction to public policy. Edward Elgar Pub-
lishing.
[71]
Michael Ramesh. 2003. Studying public policy: Policy cycles and policy subsystems.
Don Mills, Ont.: Oxford University Press.
[72]
William Riggs, Melissa Ruhl, Caroline Rodier, and Will Baumgardner. 2019. De-
signing Streets for Autonomous Vehicles. In Road Vehicle Automation 6. Springer
International Publishing, 111–122. https://doi.org/10.1007/978-3- 030-22933-7_12
[73]
Horst W. J. Rittel and Melvin M. Webber. 1973. Dilemmas in a General theory
of planning. Policy Sciences 4, 2 (June 1973), 155–169. https://doi.org/10.1007/
BF01405730
[74]
Virpi Roto, Jung-Joo Lee, Ee Lai-Chong Law, and John Zimmerman. 2021. The
overlaps and boundaries between service design and user experience design. In
Designing Interactive Systems Conference 2021. 1915–1926.
[75]
Hauke Sandhaus, WendyJu, and Qian Yang. 2023. Towards Prototyping Driverless
Vehicle Behaviors, City Design, and Policies Simultaneously. In Workshop on
Designing Technology and Policy Simultaneously: Towards A Research Agenda and
New Practice. arXiv:2304.06639 [cs.HC]
[76]
Tamara Savage. 2023. Improving Technology Forecasting by Including Policy,
Economic, and Social Factors. (9 2023). https://doi.org/10.1184/R1/24103998.v1
[77]
Anne Spaa, Abigail Durrant, Chris Elsden, and John Vines. 2019. Understand-
ing the Boundaries between Policymaking and HCI. Association for Computing
Machinery, New York, NY, USA, 1–15. https://doi.org/10.1145/3290605.3300314
[78]
Anne Spaa, Nick Spencer, Abigail Durrant, and John Vines. 2022. Creative and
collaborative reective thinking to support policy deliberation and decision
making. Evidence & Policy: A Journal of Research, Debate and Practice 18 (04 2022).
https://doi.org/10.1332/174426421X16474564583952
[79] Susan Leigh Star and Karen Ruhleder. 1994. Steps towards an ecology of infras-
tructure: complex problems in design and access for large-scale collaborative
systems. In Proceedings of the 1994 ACM conference on Computer supported coop-
erative work. 253–264.
[80]
Stephanie B Steinhardt and Steven J Jackson. 2014. Reconciling rhythms: plans
and temporal alignment in collaborative scientic work. In Proceedings of the
17th ACM conference on Computer supported cooperative work & social computing.
134–145.
[81]
Stephanie B Steinhardt and Steven J Jackson. 2015. Anticipation work: Cultivating
vision in collective practice. In Proceedings of the 18th ACM conference on computer
supported cooperative work & social computing. 443–453.
[82]
Marc Stickdorn and Jakob Schneider. 2012. This is service design thinking: Basics,
tools, cases. John Wiley & Sons.
[83]
Montreal The Global Partnership of Articial Intelligence Innova-
tion Workshop. 2023. Policy Brief: Generative AI, Jobs, and Policy Re-
sponse. Technical Report. The Global Partnership of Articial Intelligence.
https://gpai.ai/projects/future-of- work/policy-brief- generative-ai-jobs- and-
policy-response- innovation-workshop- montreal- 2023.pdf
[84]
Daniel Tkacik. 2020. CyLab researchers design privacy icon to be used by
California law. https://www.cylab.cmu.edu/news/2020/12/11-donotsell.html
[85]
Carlos Toxtli, Siddharth Suri, and Saiph Savage. 2021. Quantifying the invisible
labor in crowd work. Proceedings of the ACM on Human-Computer Interaction 5,
CSCW2 (2021), 1–26.
[86]
Lachlan Urquhart and Tom Rodden. 2017. New directions in information tech-
nology law: learning from human-computer interaction. International Review of
Law, Computers & Technology 31, 2 (2017), 150–169.
[87]
Lachlan D. Urquhart, Glenn McGarry, and Andy Crabtree. 2022. Legal Provo-
cations for HCI in the Design and Development of Trustworthy Autonomous
Systems. In Nordic Human-Computer Interaction Conference (Aarhus, Denmark)
(NordiCHI ’22). Association for Computing Machinery, New York, NY, USA, Arti-
cle 75, 12 pages. https://doi.org/10.1145/3546155.3546690
[88]
Synthetic Users. 2023. Synthetic Users: User Research without the Headaches.
https://www.syntheticusers.com/
[89]
Merlijn Van Hulst and Dvora Yanow. 2016. From policy “frames” to “framing”
theorizing a more dynamic, political approach. The American review of public
administration 46, 1 (2016), 92–112.
CHI ’24, May 11–16, 2024, Honolulu, HI, USA Yang et al.
[90]
Tom Wheeler, Niam Yaraghi Nicol Turner Lee, John Villasenor, and Nicol
Turner Lee Norman Eisen. 2023. Key enforcement issues of the AI act should
lead EU trilogue debate. https://www.brookings.edu/articles/key-enforcement-
issues-of- the-ai- act- should-lead- eu- trilogue-debate/
[91]
Cedric Deslandes Whitney, Teresa Naval, Elizabeth Quepons, Simrandeep Singh,
Steven R Rick, and Lilly Irani. 2021. HCI Tactics for Politics from Below: Meeting
the Challenges of Smart Cities. In Proceedings of the 2021 CHI Conference on
Human Factors in Computing Systems (Yokohama, Japan) (CHI ’21). Association
for Computing Machinery, New York, NY, USA, Article 297, 15 pages. https:
//doi.org/10.1145/3411764.3445314
[92]
Kyle Wiggers. 2023. The current legal cases against Generative AI are just the
beginning. https://techcrunch.com/2023/01/27/the-current- legal-cases-against-
generative-ai- are-just- the- beginning/
[93]
Richmond Y Wong and Steven J Jackson. 2015. Wireless visions: Infrastructure,
imagination, and US spectrum policy. In Proceedings of the 18th ACM Conference
on Computer Supported Cooperative Work & Social Computing. 105–115.
[94]
Richmond Y Wong and Deirdre K Mulligan. 2019. Bringing design to the privacy
table: Broadening “design” in “privacy by design” through the lens of HCI. In
Proceedings of the 2019 CHI conference on human factors in computing systems.
1–17.
[95]
Qian Yang, Aaron Steinfeld, Carolyn Rosé, and John Zimmerman. 2020. Re-
Examining Whether, Why, and How Human-AI Interaction Is Uniquely Dicult
to Design. In Proceedings of the 2020 CHI Conference on Human Factors in Comput-
ing Systems (Honolulu, HI, USA) (CHI ’20). Association for Computing Machinery,
New York, NY, USA, 1–13. https://doi.org/10.1145/3313831.3376301
[96]
Qian Yang, Richmond Y. Wong, Thomas Gilbert, Margaret D. Hagan, Steven
Jackson, Sabine Junginger, and John Zimmerman. 2023. Designing Technology and
Policy Simultaneously: Towards A Research Agenda and New Practice. Association
for Computing Machinery, New York, NY, USA. https://doi.org/10.1145/3544549.
3573827
[97]
Meg Young, Lassana Magassa, and Batya Friedman. 2019. Toward inclusive tech
policy design: a method for underrepresented voices to strengthen tech policy
documents. Ethics and Information Technology 21 (2019), 89–103.
[98]
JD Zamrescu-Pereira, Richmond Y Wong, Bjoern Hartmann, and Qian Yang.
2023. Why Johnny can’t prompt: how non-AI experts try (and fail) to design
LLM prompts. In Proceedings of the 2023 CHI Conference on Human Factors in
Computing Systems. 1–21.
APPENDIX
A DATA ANALYSIS PROCESS
We started by proposing many possible problem-solution frames
that meet the following three criteria. Next, we iteratively critiqued
and improved these emergent frames, also based on these criteria.
•
Eective problem framing: It can help explain the disparate success
levels observed in HCI-policy collaborations so far;
•
Useful solution framing: It can reveal novel strategies for address-
ing the disparity;
•
Adaptive solution framing: The strategies it reveals are exible,
allowing HCI researchers and practitioners to derive specic
actions in diverse situations. This exibility is crucial, because
eective approaches to HCI-policy collaboration vary based on
contexts [
38
]. Our goal is not to prescribe xed methods, but to
identify new avenues of improvements.
Besides meeting the above criteria, the nal frame is also a use-
ful seed for HCI-community-wide discussions. Even the authors
lack consensus on whether this problem-solution frame might be
too progressive and controversial, or too obvious that it represents
what HCI needs to do. In this sense, we see this frame as an in-
vitation (and potentially provocation) for fellow HCI researchers,
practitioners, and educators to debate the nature of HCI-policy
collaboration challenges. We encourage them to share their frames
and collectively deliberate how to best move forward.
B WHY INTERACTIVE, INFRASTRUCTURE
TECHNOLOGIES NECESSITATE DEEP,
CONTINUOUS HCI-POLICY PARTNERSHIP
AI Foundation Models (FMs) such as chatGPT are telling examples.
FMs are machine learning models that AI engineers and end users
alike can easily adapt and modify to create bespoke text-, image-,
or video-generation applications [
16
,
98
]. Without policy expertise,
HCI’s futuring and empirical methods cannot eectively understand
FMs’ societal impact or collect empirical evidence proving this
impact.
•
Each FM powers an AI-model-and-app ecosystem. HCI
cannot understand the FM’s full impact without consid-
ering the policies that govern it. For example, chatGPT. Its
societal impact depends on the various ne-tuned models and
downstream apps people build with it [
19
]. Merely empirically
studying how one model (e.g., chatGPT, August 8th, 2023 version)
impacts one user here and now misses the forest for the trees.
Instead, HCI empirical work must also understand how policies
incentivize and regulate the chatGPT-derived-model-and-app
ecosystem to understand its societal impact (the “forest.”)
This need for policy considerations is not a coincidence. From
sensor networks covering a smart city to algorithm-mediated
social media platforms, infrastructure technologies operate at
societal-level scales and power an ecosystem of various down-
stream applications. To understand their societal impact, one
must rst understand the policies that govern the ecosystem.
•
HCI cannot prove the relationship between FM design
choices and societal impact, without HCI and policy join
forces. Companies like OpenAI, which own FMs and their ecosys-
tems, hold enormous power over who can know which aspects
of FMs’ inner workings and data practices. However, this infor-
mation is critical for studying FM’s legality and social impact.
Without the joint forces of policy mandates and HCI knowledge
(e.g., on what information is critical for understanding FM’s fair-
ness), these companies may never provide the public meaningful
access to FMs’ inner workings.
This need for policy mandates on technology transparency is not
coincidental either. The eect of cyberinfrastructure on people
is often indirect and invisible. Understanding this eect requires
information about its inner workings. However, organizations
that own infrastructure technologies (e.g., social media giants,
governments) are few and powerful, holding enormous power
in guarding this information. In this context, HCI needs policy’s
power and policy expertise, which can enforce information dis-
closure and know how to do so without harming tech industry
competition.
•
Given that FMs’ societal impact is far-reaching and de-
mands quick regulation, empirical researchers need to an-
alyze their societal impact with policymakers side-by-side.
Unlike in some previous cases, HCI does not seem to have the
luxury of time to amass empirical evidence of FMs’ societal pros
and cons before policymakers demand them [41].
The fact that FMs seem to have impacted “everything everywhere
all at once” is not coincidental. From Uber (mobile computing
combined with gig work) to social media feed-ranking algorithms,
infrastructure technologies, once established and adopted at scale,
can directly impact human and human-computer interactions
of all sorts. Understanding their societal impact requires HCI
and policy to collaborate synchronously. Moreover, they need to
The Future of HCI-Policy Collaboration CHI ’24, May 11–16, 2024, Honolulu, HI, USA
develop such understanding within specic windows of opportu-
nity, for example, before the technology is widely adopted and
its associated human practices settle in.
For HCI designers who innovate human-centered systems, policy
considerations are also crucial from the outset.
•
If the FM violates the law, HCI’s novel FM apps may also
violate the law. By creating one accessible system design, HCI
incrementally improves web accessibility. Creating one new GPT
app without policy considerations, however, HCI risks violating
copyright laws.
•
FMs’ unprecedented malleability means that HCI’s well-
intentioned app designs can easily be misused. It seems
naive to think an FM tool that predicts the likely outcomes of
a user study protocol would not have been used to power “user
studies without users” [
88
], or an FM tool that helps Reddit mod-
erators to predict upcoming fake news would not have been used
to generate fake news instead. Without policy guardrails to reg-
ulate bad actors, HCI’s well-intentioned app designs can play an
unintended role in FMs’ societal harms.
These seemingly unusual risks of HCI designing illegal or unethi-
cal things are not coincidental. Just like VR platforms transformed
how governments oer public services [
7
] and GPT changed how
HCI software designers design apps, infrastructure technologies
change how almost everyone—including HCI researchers, practi-
tioners, policymakers, and policy enforcement agencies—works.
These changes complicate HCI’s eorts to improve or regulate
infrastructure technologies’ societal impact, oering yet another
reason for HCI to collaborate with policy experts.
C
RESPONDING TO REGULATORY AGENCIES’
REQUESTS FOR COMMENT
In the U.S., regulatory agencies (they are responsible for creating
rules that enact laws into practice, such as the FTC or National
Institute of Standards and Technology) continually seek input and
feedback from both experts and the general public, often through
requests for public comments for 30-60 days, allowing the public
to comment or submit data related to proposed rules or decision-
making [
1
]. HCI professionals can submit empirical evidence of
tech’s human impact as expert evidence. This evidence could inform
policies that both regulate (or promote) certain forms of technology
development and use, or that regulate (or promote certain human
behaviors related to technology use.
