2018 Article The Moral Machine

Abstract

With the rapid development of artificial intelligence have come concerns about how machines will make moral decisions, and the major challenge of quantifying societal expectations about the ethical principles that should guide machine behaviour. To address this challenge, we deployed the Moral Machine, an online experimental platform designed to explore the moral dilemmas faced by autonomous vehicles. This platform gathered 40 million decisions in ten languages from millions of people in 233 countries and territories. Here we describe the results of this experiment. First, we summarize global moral preferences. Second, we document individual variations in preferences, based on respondents’ demographics. Third, we report cross-cultural ethical variation, and uncover three major clusters of countries. Fourth, we show that these differences correlate with modern institutions and deep cultural traits. We discuss how these preferences can contribute to developing global, socially acceptable principles for machine ethics. All data used in this article are publicly available.

Full text

ARTICLE https://doi.org/10.1038/s41586-018-0637-6
The Moral Machine experiment
Edmond Awad1, Sohan Dsouza1, Richard Kim1, Jonathan Schulz2, Joseph Henrich2, Azim Shariff3*, Jean-François Bonnefon4* &
Iyad Rahwan1,5*
With the rapid development of artificial intelligence have come concerns about how machines will make moral decisions,
and the major challenge of quantifying societal expectations about the ethical principles that should guide machine
behaviour. To address this challenge, we deployed the Moral Machine, an online experimental platform designed to
explore the moral dilemmas faced by autonomous vehicles. This platform gathered 40 million decisions in ten languages
from millions of people in 233 countries and territories. Here we describe the results of this experiment. First, we
summarize global moral preferences. Second, we document individual variations in preferences, based on respondents’
demographics. Third, we report cross-cultural ethical variation, and uncover three major clusters of countries. Fourth, we
show that these differences correlate with modern institutions and deep cultural traits. We discuss how these preferences
can contribute to developing global, socially acceptable principles for machine ethics. All data used in this article are
publicly available.
We are entering an age in which machines are tasked not only to pro-
mote well-being and minimize harm, but also to distribute the well-
being they create, and the harm they cannot eliminate. Distribution
of well-being and harm inevitably creates tradeoffs, whose resolution
falls in the moral domain
1–3
. Think of an autonomous vehicle that is
about to crash, and cannot find a trajectory that would save everyone.
Should it swerve onto one jaywalking teenager to spare its three elderly
passengers? Even in the more common instances in which harm is not
inevitable, but just possible, autonomous vehicles will need to decide
how to divide up the risk of harm between the different stakeholders
on the road. Car manufacturers and policymakers are currently strug-
gling with these moral dilemmas, in large part because they cannot
be solved by any simple normative ethical principles such as Asimov’s
laws of robotics4.
Asimov’s laws were not designed to solve the problem of universal
machine ethics, and they were not even designed to let machines
distribute harm between humans. They were a narrative device whose
goal was to generate good stories, by showcasing how challenging it
is to create moral machines with a dozen lines of code. And yet, we
do not have the luxury of giving up on creating moral machines5–8.
Autonomous vehicles will cruise our roads soon, necessitating
agreement on the principles that should apply when, inevitably, life-
threatening dilemmas emerge. The frequency at which these dilemmas
will emerge is extremely hard to estimate, just as it is extremely hard to
estimate the rate at which human drivers find themselves in comparable
situations. Human drivers who die in crashes cannot report whether
they were faced with a dilemma; and human drivers who survive a
crash may not have realized that they were in a dilemma situation.
Note, though, that ethical guidelines for autonomous vehicle choices in
dilemma situations do not depend on the frequency of these situations.
Regardless of how rare these cases are, we need to agree beforehand
how they should be solved.
The key word here is ‘we’. As emphasized by former US president
Barack Obama9, consensus in this matter is going to be important.
Decisions about the ethical principles that will guide autonomous vehi-
cles cannot be left solely to either the engineers or the ethicists. For con-
sumers to switch from traditional human-driven cars to autonomous
vehicles, and for the wider public to accept the proliferation of artificial
intelligence-driven vehicles on their roads, both groups will need to
understand the origins of the ethical principles that are programmed
into these vehicles10. In other words, even if ethicists were to agree on
how autonomous vehicles should solve moral dilemmas, their work
would be useless if citizens were to disagree with their solution, and
thus opt out of the future that autonomous vehicles promise in lieu of
the status quo. Any attempt to devise artificial intelligence ethics must
be at least cognizant of public morality.
Accordingly, we need to gauge social expectations about how auton-
omous vehicles should solve moral dilemmas. This enterprise, how-
ever, is not without challenges11. The first challenge comes from the
high dimensionality of the problem. In a typical survey, one may test
whether people prefer to spare many lives rather than few9,12,13; or
whether people prefer to spare the young rather than the elderly14,15;
or whether people prefer to spare pedestrians who cross legally, rather
than pedestrians who jaywalk; or yet some other preference, or a sim-
ple combination of two or three of these preferences. But combining a
dozen such preferences leads to millions of possible scenarios, requiring
a sample size that defies any conventional method of data collection.
The second challenge makes sample size requirements even more
daunting: if we are to make progress towards universal machine ethics
(or at least to identify the obstacles thereto), we need a fine-grained under-
standing of how different individuals and countries may differ in their eth-
ical preferences
16,17
. As a result, data must be collected worldwide, in order
to assess demographic and cultural moderators of ethical preferences.
As a response to these challenges, we designed the Moral Machine,
a multilingual online ‘serious game’ for collecting large-scale data on
how citizens would want autonomous vehicles to solve moral dilemmas
in the context of unavoidable accidents. The Moral Machine attracted
worldwide attention, and allowed us to collect 39.61million decisions
from 233 countries, dependencies, or territories (Fig.1a). In the main
interface of the Moral Machine, users are shown unavoidable accident
scenarios with two possible outcomes, depending on whether the
autonomous vehicle swerves or stays on course (Fig.1b). They then
click on the outcome that they find preferable. Accident scenarios are
generated by the Moral Machine following an exploration strategy that
1The Media Lab, Massachusetts Institute of Technology, Cambridge, MA, USA. 2Department of Human Evolutionary Biology, Harvard University, Cambridge, MA, USA. 3Department of Psychology,
University of British Columbia, Vancouver, British Columbia, Canada. 4Toulouse School of Economics (TSM-R), CNRS, Université Toulouse Capitole, Toulouse, France. 5Institute for Data, Systems &
Society, Massachusetts Institute of Technology, Cambridge, MA, USA. *e-mail: shariff@psych.ubc.ca; jean-francois.bonnefon@tse-fr.eu; irahwan@mit.edu
1 NOVEMBER 2018 | VOL 563 | NATURE | 59
© 2018 Springer Nature Limited. All rights reserved.

Article
reSeArcH
focuses on nine factors: sparing humans (versus pets), staying on course
(versus swerving), sparing passengers (versus pedestrians), sparing
more lives (versus fewer lives), sparing men (versus women), sparing
the young (versus the elderly), sparing pedestrians who cross legally
(versus jaywalking), sparing the fit (versus the less fit), and sparing
those with higher social status (versus lower social status). Additional
characters were included in some scenarios (for example, criminals,
pregnant women or doctors), who were not linked to any of these nine
factors. These characters mostly served to make scenarios less repetitive
for the users. After completing a 13-accident session, participants could
complete a survey that collected, among other variables, demographic
information such as gender, age, income, and education, as well as
religious and political attitudes. Participants were geolocated so that
their coordinates could be used in a clustering analysis that sought to
identify groups of countries or territories with homogeneous vectors
of moral preferences.
Here we report the findings of the Moral Machine experiment, focus-
ing on four levels of analysis, and considering for each level of analysis
how the Moral Machine results can trace our path to universal machine
ethics. First, what are the relative importances of the nine preferences
we explored on the platform, when data are aggregated worldwide?
Second, does the intensity of each preference depend on the individual
characteristics of respondents? Third, can we identify clusters of coun-
tries with homogeneous vectors of moral preferences? And fourth, do
cultural and economic variations between countries predict variations
in their vectors of moral preferences?
Global preferences
To test the relative importance of the nine preferences simultaneously
explored by the Moral Machine, we used conjoint analysis to compute
the average marginal component effect (AMCE) of each attribute (male
character versus female character, passengers versus pedestrians, and so
on)
18
. Figure2a shows the unbiased estimates of nine AMCEs extracted
from the Moral Machine data. In each row, the bar shows the differ-
ence between the probability of sparing characters with the attribute
on the right side, and the probability of sparing the characters with the
attribute on the left side, over the joint distribution of all other attrib-
utes (seeSupplementary Information for computational details and
assumptions, and see Extended Data Figs.1, 2 for robustness checks).
As shown in Fig.2a, the strongest preferences are observed for spar
-
ing humans over animals, sparing more lives, and sparing young lives.
Accordingly, these three preferences may be considered essential build-
ing blocks for machine ethics, or at least essential topics to be consid-
ered by policymakers. Indeed, these three preferences differ starkly in
the level of controversy they are likely to raise among ethicists.
Consider, as a case in point, the ethical rules proposed in 2017 by the
German Ethics Commission on Automated and Connected Driving
19
.
This report represents the first and only attempt so far to provide offi-
cial guidelines for the ethical choices of autonomous vehicles. As such,
it provides an important context for interpreting our findings and their
relevance to other countries that might attempt to follow the German
example in the future. German Ethical Rule number 7 unambiguously
states that in dilemma situations, the protection of human life should
enjoy top priority over the protection of other animal life. This rule
is in clear agreement with social expectations assessed through the
Moral Machine. On the other hand, German Ethical Rule number 9
does not take a clear stance on whether and when autonomous vehicles
should be programmed to sacrifice the few to spare the many, but leaves
this possibility open: it is important, thus, to know that there would
be strong public agreement with such programming, even if it is not
mandated through regulation.
By contrast, German Ethical Rule number 9 also states that any dis-
tinction based on personal features, such as age, should be prohibited.
This clearly clashes with the strong preference for sparing the young
(such as children) that is assessed through the Moral Machine (see
Fig.2b for a stark illustration: the four most spared characters are the
baby, the little girl, the little boy, and the pregnant woman). This does
not mean that policymakers should necessarily go with public opinion
and allow autonomous vehicles to preferentially spare children, or,
for that matter, women over men, athletes over overweight persons,
or executives over homeless persons—for all of which we see weaker
but clear effects. But given the strong preference for sparing children,
policymakers must be aware of a dual challenge if they decide not to
give a special status to children: the challenge of explaining the rationale
for such a decision, and the challenge of handling the strong backlash
that will inevitably occur the day an autonomous vehicle sacrifices
children in a dilemma situation.
Individual variations
We assessed individual variations by further analysing the responses
of the subgroup of Moral Machine users (n=492,921) who completed
the optional demographic survey on age, education, gender, income,
and political and religious views, to assess whether preferences were
modulated by these six characteristics. First, when we include all six
characteristic variables in regression-based estimators of each of the
nine attributes, we find that individual variations have no sizable impact
on any of the nine attributes (all below 0.1; see Extended Data Table1).
Of these, the most notable effects are driven by gender and religiosity
of respondents. For example, male respondents are 0.06% less inclined
to spare females, whereas one increase in standard deviation of relig-
iosity of the respondent is associated with 0.09% more inclination to
spare humans.
More importantly, none of the six characteristics splits its sub-
populations into opposing directions of effect. On the basis of a
a
bWhat should the self-driving car do?
Fig. 1 | Coverage and interface. a, World map highlighting the locations
of Moral Machine visitors. Each point represents a location from which
at least one visitor made at least one decision (n=39.6million). The
numbers of visitors or decisions from each location are not represented.
b, Moral Machine interface. An autonomous vehicle experiences a sudden
brake failure. Staying on course would result in the death of two elderly
men and an elderly woman who are crossing on a ‘do not cross’ signal
(left). Swerving would result in the death of three passengers: an adult
man, an adult woman, and a boy (right).
60 | NATURE | VOL 563 | 1 NOVEMBER 2018
© 2018 Springer Nature Limited. All rights reserved.

Article reSeArcH
unilateral dichotomization of each of the six attributes, resulting in
two subpopulations for each, the difference in probability (ΔP) has a
positive value for all considered subpopulations. For example, both
male and female respondents indicated preference for sparing females,
but the latter group showed a stronger preference (Extended Data
Fig.3). In summary, the individual variationsthat we observe are
theoretically important, but not essential information for policymakers.
Cultural clusters
Geolocation allowed us to identify the country of residence of Moral
Machine respondents, and to seek clusters of countries with homo-
geneous vectors of moral preferences. We selected the 130 countries
with at least 100 respondents (n range101–448,125), standardized
the nine target AMCEs of each country, and conducted a hierarchical
clustering on these nine scores, using Euclidean distance and Ward’s
minimum variance method
20
. This analysis identified three distinct
‘moral clusters’ of countries. These are shown in Fig.3a, and are broadly
consistent with both geographical and cultural proximity according to
the Inglehart–Welzel Cultural Map 2010–201421.
The first cluster (which we label the Western cluster) contains North
America as well as many European countries of Protestant, Catholic,
and Orthodox Christian cultural groups. The internal structure
within this cluster also exhibits notable face validity, with a sub-cluster
containing Scandinavian countries, and a sub-cluster containing
Commonwealth countries.
The second cluster (which we call the Eastern cluster) contains
many far eastern countries such as Japan and Taiwan that belong to the
Confucianist cultural group, and Islamic countries such as Indonesia,
Pakistan and Saudi Arabia.
The third cluster (a broadly Southern cluster) consists of the Latin
American countries of Central and South America, in addition to some
countries that are characterized in part by French influence (for example,
metropolitan France, French overseas territories, and territories
that were at some point under French leadership). Latin American
countries are cleanly separated in their own sub-cluster within the
Southern cluster.
To rule out the potential effect of language, we found that the same
clusters also emerged when the clustering analysis was restricted to
participants who relied only on the pictorial representations of the
dilemmas, without accessing their written descriptions (Extended
Data Fig.4).
This clustering pattern (which is fairly robust; Extended Data Fig.5)
suggests that geographical and cultural proximity may allow groups
of territories to converge on shared preferences for machine ethics.
Between-cluster differences, though, may pose greater problems. As
shown in Fig.3b, clusters largely differ in the weight they give to some
preferences. For example, the preference to spare younger characters
rather than older characters is much less pronounced for countries
in the Eastern cluster, and much higher for countries in the Southern
cluster. The same is true for the preference for sparing higher status
characters. Similarly, countries in the Southern cluster exhibit a much
weaker preference for sparing humans over pets, compared to the other
two clusters. Only the (weak) preference for sparing pedestrians over
passengers and the (moderate) preference for sparing the lawful over
the unlawful appear to be shared to the same extent in all clusters.
Finally, we observe some striking peculiarities, such as the strong
preference for sparing women and the strong preference for sparing
fit characters in the Southern cluster. All the patterns of similarities
and differences unveiled in Fig.3b, though, suggest that manufactur-
ers and policymakers should be, if not responsive, at least cognizant
of moral preferences in the countries in which they design artificial
intelligence systems and policies. Whereas the ethical preferences of the
public should not necessarily be the primary arbiter of ethical policy,
the people’s willingness to buy autonomous vehicles and tolerate them
on the roads will depend on the palatability of the ethical rules that
are adopted.
Country-level predictors
Preferences revealed by the Moral Machine are highly correlated to
cultural and economic variations between countries. These correlations
provide support for the external validity of the platform, despite the
self-selected nature of our sample. Although we do not attempt to pin
down the ultimate reason or mechanism behind these correlations, we
document them here as they point to possible deeper explanations of
the cross-country differences and the clusters identified above.
As an illustration, consider the distance between the UnitedStates
and other countries in terms of the moral preferences extracted from
the Moral Machine (‘MM distance’). Figure4c shows a substantial
Preference for inaction
Sparing pedestrians
Sparing the lawful
Sparing females
Sparing the t
Sparing higher status
Sparing more characters
Sparing the young
Sparing humans
Preference for action
Sparing passengers
Sparing the unlawful
Sparing males
Sparing the large
Sparing lower status
Sparing fewer characters
Sparing the elderly
Sparing pets
Intervention
Relation to AV
Gender
Fitness
Social Status
Law
Age
No. character
s
Species
Preference in favour of the choice on the right side
Stroller
Girl
Boy
Pregnant
Male doctor
Female doctor
Female athlete
Executive female
Male athlete
Executive male
Large woman
Large man
Homeless
Old man
Old woman
Dog
Criminal
Cat
Prefer
ence in favour of sparing charactersa b
No change +0.2 +0.4 +0.6 +0.8
6
P
1234
–0.2 –0.1 +0.2+0.1No
change
Fig. 2 | Global preferences. a, AMCE for each preference. In each row, ΔP
is the difference between the probability of sparing characters possessing
the attribute on the right, and the probability of sparing characters
possessing the attribute on the left, aggregated over all other attributes.
For example, for the attribute age, the probability of sparing young
characters is 0.49 (s.e. =0.0008) greater than the probability of sparing
older characters. The 95% confidence intervals of the means are omitted
owing to their insignificant width, given the sample size (n=35.2million).
For the number of characters (No. characters), effect sizes are shown
for each number of additional characters (1 to 4; n1=1.52million,
n2=1.52million, n3=1.52million, n4=1.53million); the effect size for
two additional characters overlaps with the mean effect of the attribute.AV,
autonomous vehicle. b, Relative advantage or penalty for each character,
compared to an adult man or woman. For each character, ΔP is the
difference the between the probability of sparing this character (when
presented alone) and the probability of sparing one adult man or woman
(n=1million). For example, the probability of sparing a girl is 0.15 (s.e.
=0.003) higher than the probability of sparing an adult man or woman.
1 NOVEMBER 2018 | VOL 563 | NATURE | 61
© 2018 Springer Nature Limited. All rights reserved.

ArticlereSeArcH
correlation (ρ=0.49) between this MM distance and the cultural
distance from the UnitedStates based on the World Values Survey
22
.
In other words, the more culturally similar a country is to the
UnitedStates, the more similarly its people play the Moral Machine.
Next, we highlight four important cultural and economic predic-
tors of Moral Machine preferences. First, we observe systematic dif-
ferences between individualistic cultures and collectivistic cultures23.
Participants from individualistic cultures, which emphasize the distinc-
tive value of each individual
23
, show a stronger preference for sparing
the greater number of characters (Fig.4a). Furthermore, participants
from collectivistic cultures, which emphasize the respect that is due
to older members of the community
23
, show a weaker preference for
sparing younger characters (Fig.4a, inset). Because the preference for
sparing the many and the preference for sparing the young are arguably
the most important for policymakers to consider, this split between
individualistic and collectivistic cultures may prove an important obsta-
cle for universal machine ethics (seeSupplementary Information).
Another important (yet under-discussed) question for policymakers
to consider is the importance of whether pedestrians are abiding by
or violating the law. Should those who are crossing the street illegally
benefit from the same protection as pedestrians who cross legally? Or
should the primacy of their protection in comparison to other ethical
priorities be reduced? We observe that prosperity (as indexed by GDP
per capita
24
) and the quality of rules and institutions (as indexed by
the Rule of Law
25
) correlate with a greater preference against pedes-
trians who cross illegally (Fig.4b). In other words, participants from
countries that are poorer and suffer from weaker institutions are more
tolerant of pedestrians who cross illegally, presumably because of their
experience of lower rule compliance and weaker punishment of rule
deviation
26
. This observation limits the generalizability of the recent
German ethics guideline, for example, which state that “parties involved
in the generation of mobility risks must not sacrifice non-involved
parties.” (seeSupplementary Information).
Finally, our data reveal a set of preferences in which certain char-
acters are preferred for demographic reasons. First, we observe that
higher country-level economic inequality (as indexed by the country’s
Gini coefficient) corresponds to how unequally characters of different
social status are treated. Those from countries with less economic
equality between the rich and poor also treat the rich and poor less
equally in the Moral Machine. This relationship may be explained by
regular encounters with inequality seeping into people’s moral prefer-
ences, or perhaps because broader egalitarian norms affect both how
Maldives
Azerbaijan
Turkey
Honduras
Guatemala
Bahamas
Martinique
New Caledonia
Reunion
Oman
Madagascar
Angola
Peru
El Salvador
Paraguay
Guernsey
Iran
Nepal
Pakistan
Andorra
Cambodia
Japan
Macao
China
Brunei
Monaco
Afghanistan
Argentina
Uruguay
Bolivia
Ecuador
Colombia
Venezuela
Panama
Philippines
Chile
Malta
Mongolia
French Polynesia
Jordan
Palestinian Territory
United Arab Emirates
South Korea
Brazil
Guam
Bangladesh
Jersey
South Africa
Isle of Man
Myanmar
Puerto Rico
Algeria
Morocco
Dominican Republic
France
Czechia
Armenia
Macedonia
India
Mauritius
Egypt
Lebanon
Taiwan
Thailand
Bahrain
Hong Kong
Kazakhstan
Iraq
Israel
Barbados
Lithuania
Nicaragua
Syria
Uzbekistan
Nigeria
Singapore
Sri Lanka
Costa Rica
Mexico
Hungary
Slovakia
Kuwait
Saudi Arabia
Indonesia
Malaysia
Georgia
Moldova
Belarus
Slovenia
Jamaica
Trinidad and Tobago
Albania
Qatar
Tunisia
Vietnam
Canada
United States
Ireland
Kenya
Kyrgyzstan
Bosnia and
Herzegovina
Greece
Serbia
Croatia
Russia
Ukraine
Latvia
Poland
New Zealand
Portugal
Spain
Belgium
Montenegro
Estonia
Romania
Bulgaria
Italy
Cyprus
Sweden
Iceland
Finland
The Netherlands
Australia
United Kingdom
Austria
Luxembourg
Denmark
Norway
Germany
Switzerland
z = 0
z = 1
z = 2
z = 0
z = 1
z = 2
Preference
for inaction
Sparing
pedestrians
Sparing
females
Sparing
the t
Sparing
the lawful
Sparing
higher statu
s
Sparing
the young
Sparing
more
Sparing
humans
z = 0
z = 1
z = 2
Eastern
Western Southern
a
b
Eastern
Western
Southern
South Asi
a
Protestant
Orthodox
Latin
America
Islamic
English
Confucian
Catholic
Baltic
Other
Western
Eastern
Southern
1.0
0.75
0.5
0.25
0.0
Preference
for inaction
Sparing
pedestrians
Sparing
females
Sparing
the t
Sparing
the lawful
Sparing
higher statu
s
Sparing
the young
Sparing
more
Sparing
humans
Preference
for inaction
Sparing
pedestrians
Sparing
females
Sparing
the t
Sparing
the lawful
Sparing
higher statu
s
Sparing
the young
Sparing
more
Sparing
humans
Fig. 3 | Country-level clusters. a, Hierarchical cluster of countries based
on average marginal causal effect. One hundred and thirty countries with
at least 100 respondents were selected (range,101–448,125). The three
colours of the dendrogram branches represent three large clusters—Western,
Eastern, and Southern. Country names are coloured according to the
Inglehart–Welzel Cultural Map 2010–201421. Distributions across the three
clusters reveal stark differences. For instance, cluster 2 (Eastern) consists
mostly of countries of Islamic and Confucian cultures. By contrast, cluster
1 (Western) has large percentages of Protestant, Catholic, and Orthodox
countries in Europe. b, Mean AMCE z-scores of the three major clusters.
Radar plot of the mean AMCE z-scores of three clusters reveals a striking
pattern of differences between the clusters along the nine attributes.
For example, countries belonging to the Southern cluster show a strong
preference for sparing females compared to countries in other clusters.
62 | NATURE | VOL 563 | 1 NOVEMBER 2018
© 2018 Springer Nature Limited. All rights reserved.

Article reSeArcH
much inequality a country is willing to tolerate at the societal level,
and how much inequality participants endorse in their Moral Machine
judgments. Second, the differential treatment of male and female
characters in the Moral Machine corresponded to the country-level
gender gap in health and survival (a composite in which higher scores
indicated higher ratios of female to male life expectancy and sex ratio
at birth—a marker of female infanticide and anti-female sex-selective
abortion). In nearly all countries, participants showed a preference for
female characters; however, this preference was stronger in nations
with better health and survival prospects for women. In other words,
in places where there is less devaluation of women’s lives in health
and at birth, males are seen as more expendable in Moral Machine
decision-making (Fig.4e). While not aiming to pin down the causes
of this variation in Extended Data Table2, we nevertheless provide
a regression analysis that demonstrates that the results hold when
controlling for several potentially confounding factors.
Discussion
Never in the history of humanity have we allowed a machine to
autonomously decide who should live and who should die, in a
fraction of a second, without real-time supervision. We are going to
cross that bridge any time now, and it will not happen in a distant theatre
of military operations; it will happen in that most mundane aspect of
our lives, everyday transportation. Before we allow our cars to make
ethical decisions, we need to have a global conversation to express our
preferences to the companies that will design moral algorithms, and to
the policymakers that will regulate them.
The Moral Machine was deployed to initiate such a conversation,
and millions of people weighed in from around the world. Respondents
could be as parsimonious or thorough as they wished in the ethical
framework they decided to follow. They could engage in a complicated
weighting of all nine variables used in the Moral Machine, or adopt
simple rules such as ‘let the car always go onward’. Our data helped us
to identify three strong preferences that can serve as building blocks for
discussions of universal machine ethics, even if they are not ultimately
endorsed by policymakers: the preference for sparing human lives, the
preference for sparing more lives, and the preference for sparing young
lives. Some preferences based on gender or social status vary consid-
erably across countries, and appear to reflect underlying societal-level
preferences for egalitarianism27.
The Moral Machine project was atypical in many respects. It was
atypical in its objectives and ambitions: no research has previously
attempted to measure moral preferences using a nine-dimensional
experimental design in more than 200 countries. To achieve this
unusual objective, we deployed a viral online platform, hoping that
we would reach out to vast numbers of participants. This allowed us
to collect data from millions of people over the entire world, a feat
that would be nearly impossibly hard and costly to achieve through
standard academic survey methods. For example, recruiting nationally
representative samples of participants in hundreds of countries would
already be extremely difficult, but testing a nine-factorial design in each
of these samples would verge on impossible. Our approach allowed us
to bypass these difficulties, but its downside is that our sample is self-
selected, and not guaranteed to exactly match the socio-demographics
of each country (Extended Data Fig.6). The fact that the cross-societal
variation we observed aligns with previously established cultural clus-
ters, as well as the fact that macro-economic variables are predictive
of Moral Machine responses, are good signals about the reliability of
our data, as is our post-stratification analysis (Extended Data Fig.7
and Supplementary Information). But the fact that our samples are
not guaranteed to be representative means that policymakers should
not embrace our data as the final word on societal preferences—even
if our sample is arguably close to the internet-connected, tech-savvy
population that is interested in driverless car technology, and more
likely to participate in early adoption.
Even with a sample size as large as ours, we could not do justice to
all of the complexity of autonomous vehicle dilemmas. For example,
we did not introduce uncertainty about the fates of the characters, and
we did not introduce any uncertainty about the classification of these
characters. In our scenarios, characters were recognized as adults,
children, and so on with 100% certainty, and life-and-death outcomes
were predicted with 100% certainty. These assumptions are technolog-
ically unrealistic, but they were necessary to keep the project tractable.
Similarly, we did not manipulate the hypothetical relationship between
respondents and characters (for example, relatives or spouses). Our
previous work did not find a strong effect of this variable on moral
preferences12.
Indeed, we can embrace the challenges of machine ethics as a
unique opportunity to decide, as a community, what we believe to
be right or wrong; and to make sure that machines, unlike humans,
unerringly follow these moral preferences. We might not reach uni-
versal agreement: even the strongest preferences expressed through
the Moral Machine showed substantial cultural variations, and our
project builds on a long tradition of investigating cultural variations
in ethical judgments
28
. But the fact that broad regions of the world
displayed relative agreement suggests that our journey to consensual
machine ethics is not doomed from the start. Attempts at establishing
broad ethical codes for intelligent machines, such as the Asilomar
AI Principles29, often recommend that machine ethics should be
aligned with human values. These codes seldom recognize, though,
that humans experience inner conflict, interpersonal disagreements,
and cultural dissimilarities in the moral domain
30–32
. We have shown
that these conflicts, disagreements, and dissimilarities, while substan-
tial, may not be fatal.
Online content
Any methods, additional references, Nature Research reporting summaries, source
data, statements of data avai lability and associated accession codes are available at
https://doi.org/10.1038/s41586-018-0637-6.
0.6
0.5
0.4
Sparing more characters
a
0255075 100
Individualism
b
Sparing the lawful
0.4
0.3
0.2
0.1
–2 –1 012
Rule of law
Cluster 1 (Western)Cluster 2 (Eastern)Cluster 3 (Southern)
MM distance from US
6
4
2
0
Cultural distance from US
0.00 0.05 0.10 0.15 0.20 0.25 30 40 50 60
Economic inequality (Gini)
Sparing higher status
0.5
0.4
0.3
0.2
Sparing females
0.92 0.94 0.96
0.98
0.3
0.2
0.1
0.0
0.6
0.5
0.4
25 50 75
Individualism
Sparing the young
0.40
Sparing the lawful
0.45
0.35
0.30
0.25 3.5 4.0 4.5 5.0
log GDP pc
Gender gap in health and survival
cd e
U= 0.5
P= 1 × 10–4
U= 0.3
P= 8 × 10–4
U= 0.49
P= 1 × 10–4
U= 0.41
P= 1 × 10–4
U= 0.29
P= 0.0025
U= 0.43
P= 1 × 10–4
U= 0.34
P= 2 × 10–4
Fig. 4 | Association between Moral Machine preferences and other
variables at the country level. Each panel shows Spearman’s ρ and
P value for the correlation test between the relevant pair of variables.
a, Association between individualism and the preference for sparing
more characters (n=87), or the preference for sparing the young (inset;
n=87). b, Association between the preference for sparing the lawful and
each of rule of law (n=122) and log GDP per capita (pc) (inset; n=110).
c, Association between cultural distance from the UnitedStates and MM
distance (distance in terms of the moral preferences extracted from the
Moral Machine) from the UnitedStates (n=72). d, Association between
economic inequality (Gini coefficient) and the preference for sparing
higher status (n=98). e, Association between the gender gap in health and
survival and the preference for sparing females (n=104).
1 NOVEMBER 2018 | VOL 563 | NATURE | 63
© 2018 Springer Nature Limited. All rights reserved.

ArticlereSeArcH
Received: 2 March 2018; Accepted: 25 September 2018;
Published online 24 October 2018.
1. Greene, J. Moral Tribes: Emotion, Reason and the Gap Between Us and Them
(Atlantic Books, London, 2013).
2. Tomasello, M. A Natural History of Human Thinking(Harvard Univ. Press,
Cambridge, 2014).
3. Cushman, F. & Young, L. The psychology of dilemmas and the philosophy of
morality. Ethical Theory Moral Pract. 12, 9–24 (2009).
4. Asimov, I. I, Robot (Doubleday, New York, 1950).
5. Bryson, J. & Wineld, A. Standardizing ethical design for articial intelligence
and autonomous systems. Computer 50, 116–119 (2017).
6. Wiener, N. Some moral and technical consequences of automation. Science
131, 1355–1358 (1960).
7. Wallach, W. & Allen, C. Moral Machines: Teaching Robots Right from Wrong
(Oxford Univ. Press, Oxford, 2008).
8. Dignum, V. Responsible autonomy. InProc. 26th International Joint Conference
on Articial Intelligence 4698–4704 (IJCAI, 2017).
9. Dadich, S. Barack Obama, neural nets, self-driving cars, and the future of
the world. Wired https://www.wired.com/2016/10/president-obama-mit-joi-
ito-interview/ (2016).
10. Shari, A., Bonnefon, J.-F. & Rahwan, I. Psychological roadblocks to the adoption
of self-driving vehicles. Nat. Hum. Behav. 1, 694–696 (2017).
11. Conitzer, V., Brill, M. & Freeman, R. Crowdsourcing societal tradeos. In Proc.
2015 International Conference on Autonomous Agents and Multiagent Systems
1213–1217 (IFAAMAS, 2015).
12. Bonnefon, J.-F., Shari, A. & Rahwan, I. The social dilemma of autonomous
vehicles. Science 352, 1573–1576 (2016).
13. Hauser, M., Cushman, F., Young, L., Jin, K.-X. R. & Mikhail, J. A dissociation
between moral judgments and justications. Mind Lang. 22, 1–21 (2007).
14. Carlsson, F., Daruvala, D. & Jaldell, H. Preferences for lives, injuries, and age:
a stated preference survey. Accid. Anal. Prev. 42, 1814–1821 (2010).
15. Johansson-Stenman, O. & Martinsson, P. Are some lives more valuable? An
ethical preferences approach. J. Health Econ. 27, 739–752 (2008).
16. Johansson-Stenman, O., Mahmud, M. & Martinsson, P. Saving lives versus
life-years in rural Bangladesh: an ethical preferences approach. Health Econ.
20, 723–736 (2011).
17. Graham, J., Meindl, P., Beall, E., Johnson, K. M. & Zhang, L. Cultural dierences in
moral judgment and behavior, across and within societies. Curr. Opin. Psychol. 8,
125–130 (2016).
18. Hainmueller, J., Hopkins, D. J. & Yamamoto, T. Causal inference in conjoint
analysis: understanding multidimensional choices via stated preference
experiments. Polit. Anal. 22, 1–30 (2014).
19. Luetge, C. The German Ethics Code for automated and connected driving.
Philos. Technol. 30, 547–558 (2017).
20. Müllner, D. Modern hierarchical, agglomerative clustering algorithms. Preprint
at https://arxiv.org/abs/1109.2378 (2011).
21. Inglehart, R. & Welzel, C. Modernization, Cultural Change, and Democracy: The
Human Development Sequence (Cambridge Univ. Press, Cambridge, 2005).
22. Muthukrishna, M. Beyond WEIRD psychology: measuring and mapping scales
of cultural and psychological distance. Preprint at https://ssrn.com/
abstract=3259613 (2018).
23. Hofstede, G. Culture’s Consequences: Comparing Values, Behaviors, Institutions
and Organizations Across Nations (Sage, Thousand Oaks, 2003).
24. International Monetary Fund. World Economic Outlook Database https://
www.imf.org/external/pubs/ft/weo/2017/01/weodata/index.aspx
(2017).
25. Kaufmann, D., Kraay, A. & Mastruzzi, M. The worldwide governance
indicators: methodology and analytical issues. Hague J. Rule Law 3, 220–246
(2011).
26. Gächter, S. & Schulz, J. F. Intrinsic honesty and the prevalence of rule violations
across societies. Nature 531, 496–499 (2016).
27. O’Neil, C. Weapons of Math Destruction: How Big Data Increases Inequality and
Threatens Democracy (Penguin, London, 2016).
28. Henrich, J. et al. In search of Homo Economicus: behavioral experiments in 15
small-scale societies. Am. Econ. Rev. 91, 73–78 (2001).
29. Future of Life Institute. Asilomar AI Principles https://futureoife.org/
ai-principles/ (2017).
30. Haidt, J. The Righteous Mind: Why Good People Are Divided by Politics and
Religion (Knopf Doubleday, New York, 2012).
31. Gastil, J., Braman, D., Kahan, D. & Slovic, P. The cultural orientation of mass
political opinion. PS Polit. Sci. Polit. 44, 711–714 (2011).
32. Nishi, A., Christakis, N. A. & Rand, D. G. Cooperation, decision time, and culture:
online experiments with American and Indian participants. PLoS ONE 12,
e0171252 (2017).
Acknowledgements I.R., E.A., S.D., and R.K. acknowledge support from the
Ethics and Governance of Artificial Intelligence Fund. J.-F.B. acknowledges
support from the ANR-Labex Institute for Advanced Study in Toulouse.
Author contributions I.R., A.S. and J.-F.B. planned the research. I.R., A.S., J.-F.B.,
E.A. and S.D. designed the experiment. E.A. and S.D. built the platform and
collected the data. E.A., S.D., R.K., J.S. and A.S. analysed the data. E.A., S.D., R.K.,
J.S., J.H., A.S., J.-F.B., and I.R interpreted the results and wrote the paper.
Competing interests The authors declare no competing interests.
Additional information
Extended data is available for this paper at https://doi.org/10.1038/s41586-
018-0637-6.
Supplementary information is available for this paper at https://doi.org/
10.1038/s41586-018-0637-6.
Reprints and permissions information is available at http://www.nature.com/
reprints.
Correspondence and requests for materials should be addressed to A.S. or
J.-F.B. and I.R.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
64 | NATURE | VOL 563 | 1 NOVEMBER 2018
© 2018 Springer Nature Limited. All rights reserved.

Article reSeArcH
METHODS
This study was approved by the Institute Review Board (IRB) at Massachusetts
Institute of Technology (MIT). The authors complied with all relevant ethical con-
siderations. No statistical methods were used to predetermine sample size. The
experiments were randomized and the investigators were blinded to allocation
during experiments and outcome assessment.
The Moral Machine website was designed to col lect data on the moral accepta-
bility of decisions made by autonomous vehicles in situations of unavoidable
accidents, in which they must decide who is spared and who is sacrificed. The
Moral Machine was deployed in June 2016. In October 2016, a feature was added
that offered users the option to fill a survey about their demographics, political
views, and religious beliefs. Between November 2016 and March 2017, the website
was progressively translated into nine languages in addition to English (Arabic,
Chinese, French, German, Japanese, Korean, Portuguese, Russian, and Spanish).
While the Moral Machine offers four different modes (seeSupplementary
Information), the focus of this article is on the central data-gathering feature of
the website, called the Judge mode. In this mode, users are presented with a series
of dilemmas in which the autonomous vehicle must decide between two different
outcomes. In each dilemma, one outcome amounts to sparing a group of 1 to 5
characters (chosen from a sample of 20 characters, Fig.2b) and killing another
group of 1 to 5 characters. The other outcome reverses the fates of the two groups.
The only task of the user is to choose between the two outcomes, as a response to
the question “What should the self-driving car do?” Users have the option to click
on a button labelled ‘see description’ to display a complete text description of the
characters in the two groups, together with their fate in each outcome.
While users can go through as many dilemmas as they wish, dilemmas are
generated in sessions of 13. Within each session, one dilemma is entirely random.
The other 12 dilemmas are sampled from a space of approximately 26 million
possibilities (see below). Accordingly, it is extremely improbable for a given user
to see the same dilemma twice, regardless of how many dilemmas they choose to
go through, or how many times they visit the Moral Machine.
Leaving aside the one entirely random di lemma, there are two dilemmas within
each session that focus on each of six dimensions of moral preferences: character
gender, character age, character physical fitness, character social status, charac-
ter species, and character number. Furthermore, each dilemma simultaneously
randomizes three additional attributes: which group of characters will be spared
if the car does nothing; whether the two groups are pedestrians, or whether one
group is in the car; and whether the pedestrian characters are crossing legally or
illegally. This explorat ion strategy is supported by a dilemma generation algorithm
(seeSupplementary Information, which also provides extensive descriptions of
statistical analyses, robustness checks, and tests of internal and external validity).
After completing a session of 13 dilemmas, users are presented with a summary
of their decisions: which character they spared the most; which character they sac-
rificed the most; and the relative importance of the nine target moral dimensions
in their decisions, compared to their importance to the average of all other users so
far. Users have the option to share this summary with their social network. Either
before or after they see this summary (randomized order), users are asked whether
they want to “help us better understand their decisions.” Users who click ‘yes’ are
directed to a survey of their demographic, political, and religious characteristics.
They also have the option to edit the summary of their decisions, to tell us about
the self-perceived importance of the nine dimensions in their decisions. These
self-perceptions were not analysed in this article.
The country from which users access the website is geo-localized through
the IP address of their computer or mobile device. This information is used to
compute a vector of moral preferences for each country. In turn, these moral
vectors are used both for cultural clustering, and for country-level correlations
between moral preferences and socio-economic indicators. The source and period
of reference for each socio-economic indicator are detailed in theSupplementary
Information.
Data availability
Source data and code that can be used to reproduce Figs.2–4, Extended Data
Figs.1–7, Extended Data Tables1, 2, Supplementary Figs.3–21, and Supplementary
Table2 are all available at the following link: https://goo.gl/JXRrBP. The provided
data, both at the individual level (anonymized IDs) and the country level, can be
used beyond replication to answer follow-up research questions.
© 2018 Springer Nature Limited. All rights reserved.

ArticlereSeArcH
Extended Data Fig. 1 | Robustness checks: internal validation of three
simplifying assumptions. Calculated values correspond to values in
Fig.2a (that is, AMCE calculated using conjoint analysis). For example,
‘Sparing Pedestrians [Relation to AV]’ refers to the difference between
the probability of sparing pedestrians, and the probability of sparing
passengers (attribute name: Relation to AV), aggregated over all other
attributes. Error bars represent 95% confidence intervals of the means.
AV, autonomous vehicle. a, Validation of assumption 1 (stability and
no-carryover effect): potential outcomes remain stable regardless of
scenario order. b, Validation of assumption 2 (no profile-order effects):
potential outcomes remain stable regardless of left–right positioning of
choice options on the screen. c, Validation of assumption 3 (randomization
of the profiles): potential outcomes are statistically independent of the
profiles. This assumption should be satisfied by design. However, a
mismatch between the design and the collected data can happen during
data collection. This panel shows that using theoretical proportions (by
design) and actual proportions (in collected data) of subgroups results in
similar effect estimates. SeeSupplementary Information for more details.
© 2018 Springer Nature Limited. All rights reserved.

Article reSeArcH
Extended Data Fig. 2 | See next page for caption.
© 2018 Springer Nature Limited. All rights reserved.

ArticlereSeArcH
Extended Data Fig. 2 | Robustness checks: external validation of
three factors. Calculated values correspond to values in Fig.2a (AMCE
calculated using conjoint analysis). For example, ‘Sparing Pedestrians
[Relation to AV]’ refers to the difference between the probability of
sparing pedestrians, and the probability of sparing passengers (attribute
name: Relation to AV), aggregated over all other attributes. Error bars
represent 95% confidence intervals of the means. a, Validation of textual
description (seen versus not seen). By default, respondents see only
the visual representation of a scenario. Interpretation of what type of
characters they represent (for example, female doctor) may not be obvious.
Optionally, respondents can read a textual description of the scenario by
clicking on ‘see description’. This panel shows that direction and (except
in one case) order of effect estimates remain stable. The magnitude of the
effects increases for respondents who read the textual descriptions, which
means that the effects reported in Fig.2a were not overestimated because
of visual ambiguity. b, Validation of device used (desktop versus mobile).
Direction and order of effect estimates remain stable regardless of whether
respondents used desktop or mobile devices when completing the task.
c, Validation of data set (all data versus full first-session data versus
survey-only data). Direction and order of effect estimates remain stable
regardless of whether the data used in analysis are all data, data restricted
to only first completed (13-scenario) session by any user, or data restricted
to completed sessions after which the demographic survey was taken. First
completed session by any user is an interesting subset of the data because
respondents had not seen their summary of results yet, and respondents
ended up completing the session. Survey-only data are also interesting
given that the conclusions about individual variations in the main paper
and from Extended Data Fig.3 and Extended Data Table1 are drawn from
this subset. SeeSupplementary Information for more details.
© 2018 Springer Nature Limited. All rights reserved.

Article reSeArcH
Extended Data Fig. 3 | Average marginal causal effect (AMCE) of
attributes for different subpopulations. Subpopulations are characterized
by respondents’ age (a, older versus younger), gender (b, male versus
female), education (c, less versus more educated), income (d, higher versus
lower income), political views (e, conservative versus progressive), and
religious views (f, not religious versus very religious). Error bars represent
95% confidence intervals of the means. Note that AMCE has a positive
value for all considered subpopulations; for example, both male and female
respondents indicated a preference for sparing females, but the latter
group showed a stronger preference. SeeSupplementary Information for a
detailed description of the cutoffs and the groupings of ordinal categories
that were used to define each subpopulation.
© 2018 Springer Nature Limited. All rights reserved.

ArticlereSeArcH
Extended Data Fig. 4 | Hierarchical cluster of countries based on
country-level effect sizes calculated after filtering out responses
for which the linguistic description was seen, thus neutralizing any
potential effect of language. The three colours of the dendrogram
branches represent three large clusters: Western, Eastern, and Southern.
The names of the countries are coloured according to the Inglehart–Welzel
Cultural Map 2010–201421. SeeSupplementary Information for more
details. The dendrogram is essentially similar to that shown in Fig.3a.
© 2018 Springer Nature Limited. All rights reserved.

Article reSeArcH
Extended Data Fig. 5 | Validation of hierarchical cluster of countries.
a, b, We use two internal metrics of validation of three linkage criteria
of calculating hierarchical clustering (Ward, Complete and Average) in
addition to the K-means algorithm: a, Calinski–Harabasz index;
b, silhouette index. The xaxis indicates the number of clusters. For both
internal metrics, a higher index value indicates a ‘better’ fit of partition
to the data. c, d, We use two external metrics of validation of the used
hierarchical clustering algorithm (Ward) versus those of random
clustering assignment: c, purity; d, maximum matching. The histogram
shows the distributions of purity and maximum matching values derived
from randomly assigning countries to nine clusters. The red dotted
lines indicate purity and maximum matching values computed from the
clustering output of the hierarchical clustering algorithm using ACME
values. SeeSupplementary Information for more details.
© 2018 Springer Nature Limited. All rights reserved.

ArticlereSeArcH
Extended Data Fig. 6 | Demographic distributions of sample of
population that completed the survey on Moral Machine (MM)website.
Distributions are based on gender (a), age (b), income (c), and education
attributes (d). Most users on Moral Machine are male, went through
college, and are in their 20s or 30s. While this indicates that the users of
Moral Machine are not a representative sample of the whole population, it
is important to note that this sample at least covers broad demographics.
SeeSupplementary Information for more details.
© 2018 Springer Nature Limited. All rights reserved.

Article reSeArcH
Extended Data Fig. 7 | Demographic distributions of US sample of
population that completed the survey on Moral Machine website versus
US sample of population in American Community Survey (ACS)
data set. a–d, Only gender (a), age (b), income (c), and education (d)
attributes are available for both data sets. The MM US sample has an over-
representation of males and younger individuals compared to the ACS US
sample. e, A comparison of effect sizes as calculated for US respondents
who took the survey on MM with the use of post-stratification to match
the corresponding proportions for the ACS sample. Except for ‘Relation to
AV’ (the second smallest effect), the direction and order of all effects are
unaffected. SeeSupplementary Information for more details.
© 2018 Springer Nature Limited. All rights reserved.

ArticlereSeArcH
Extended Data Table 1 | Regression table showing the individual variations for each of the nine attributes
Dependent variables are recorded as to whether the preferred option was chosen (for example, whether the respondent spared females). Continuous predictor variables are all standardized. All
models include structural covariates (remaining attributes of a scenario). Coecients are estimated using a regression-based estimator with cluster-robust standard errors. *P<0.01, **P<0.001,
***P<0.0001. SeeSupplementary Information for more details.
© 2018 Springer Nature Limited. All rights reserved.

Article reSeArcH
Extended Data Table 2 | Country-level OLS regressions showing the relationships between key ethical preferences and various social,
political and economic measures
Pairwise exclusion was used for missing data. Predicted relationships are shown in bold. *P<0.10, **P<0.05, ***P<0.01. SeeSupplementary Information for more details.
© 2018 Springer Nature Limited. All rights reserved.

1
nature research | reporting summary April 2018
Corresponding author(s): Iyad Rahwan
Reporting Summary
Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Policy Checklist.
Statistical parameters
When statistical analyses are reported, confirm that the following items are present in the relevant location (e.g. figure legend, table legend, main
text, or Methods section).
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
An indication of whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested
A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistics including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND
variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)
For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Clearly defined error bars
State explicitly what error bars represent (e.g. SD, SE, CI)
Our web collection on statistics for biologists may be useful.
Software and code
Policy information about availability of computer code
Data collection Data was collected through the Moral Machine website (www.moralmachine.mit.edu) which was built especially for the purpose of this
study. Some data at the level of countries was collected from other work (all cited).
Data analysis Data was preprocessed and analyzed using Python (Jupyter 3.0), R (RStudio 3.4.1), and SPSS.
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers
upon request. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.
Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A list of figures that have associated raw data
- A description of any restrictions on data availability
Source data and code that can be used to reproduce Figs. 2-4; Extended Data Figs. 1-9; Supplementary Figures S3-S21; and Supplementary Table S2 are all available

2
nature research | reporting summary April 2018
at the following link: https://goo.gl/JXRrBP. The provided data, both at the individual level (anonymized IDs) and the country level, can be used beyond replication to
answer follow up research questions.
Field-specific reporting
Please select the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/authors/policies/ReportingSummary-flat.pdf
Behavioural & social sciences study design
All studies must disclose on these points even when the disclosure is negative.
Study description This study uses quantitative data collected from an online website designed as a randomized controlled multi-factorial design
experiment.
Research sample Research sample is Internet users who chose to visit the website and contribute to the data. Research sample has 2.3M users. The
demographic information is only available for 586K users (aged between 15-75; 27% are females and 2% are "others"). Detailed info and
figures about demographics of users are included in the SI document.
Sampling strategy The sample is self-selected. There was no power calculation. Sample size is 2.3M participants. For the country-level data, as a rule of
thumb, a country is included in the analysis if we have at least 100 participants from it (sample size range: [101 - 448,125]).
Data collection Data was collected via the website http://moralmachine.mit.edu. Allocations of users to conditions was done automatically by the
website, and so all researchers were blind to the experimental conditions. Data was stored in a MongoDB data base on a remote server.
Timing Website was deployed on June 23rd, 2016. The data used in the analysis was collected continuously up until Dec 20th, 2017
Data exclusions The main analysis excluded responses for which the participant took more than 30 min. This resulted in the exclusion of 33,838
responses (out of 39.6M). This criteria was established after collecting the data. The time was chosen arbitrarily but choosing more or
less time has inconsequential effect on the results. Additionally, robustness checks were performed in which we show that neither the
exclusion of unfinished sessions (sessions in which strictly less than 13 tasks were answered), nor the exclusion of responses after the
end of the first session by the same respondent have any consequential effect on results. Analysis that relates to age of participants
excluded responses for which the indicated age falls outside the range (15-75).
Non-participation Out of 2.86M completed sessions (13 scenarios), 43,979 sessions were opted out for. The number of participants dropping out (out of
2.3M) is hard to exactly know, but is comparable to the number of dropped out sessions. We did not collect the reasons to drop out.
Randomization Users who visit the website get presented with 13 scenarios that are dawn from six main different conditions (2 scenarios from each
condition + 1 fully random scenario). The six conditions vary the following aspect of characters: age, gender, fitness level, social status,
number, and whether they are humans or pets. In conjunction with these six conditions, three main conditions were randomized:
interventionism, relation to AV, and legality. Within each main condition, characters are sampled from a set of 20 characters (e.g. adult
male, female athlete, homeless person, etc.).
Reporting for specific materials, systems and methods
Materials & experimental systems
n/a Involved in the study
Unique biological materials
Antibodies
Eukaryotic cell lines
Palaeontology
Animals and other organisms
Human research participants
Methods
n/a Involved in the study
ChIP-seq
Flow cytometry
MRI-based neuroimaging

3
nature research | reporting summary April 2018
Human research participants
Policy information about studies involving human research participants
Population characteristics See above
Recruitment Research sample is Internet users who chose to visit the website and contribute to the data. Thus, the sample is self-selected,
and it is representative of a subset of the full population.