Nils Köbis
Hi!
I'm Nils, a Professor for the Human Understanding of Algorithms and Machines at the Research Center Trustworthy Data Science and Security, University Duisburg-Essen, and an affiliated researcher at the Center for Humans and Machines (Max Planck Institute for Human Development)
My work deals with corruption, (un-)ethical behavior, social norms, and, more recently, artificial intelligence.
I am a co-founder of the Interdisciplinary Corruption Research Network and, together with Matthew Stephenson and Christopher Starke founded and co-hosted the KickBack - Global AntiCorruption Podcast.
Previously, I completed a Post-Doc CREED, Department of Economics, University of Amsterdam and a Ph.D. in Social Psychology at the VU Free University Amsterdam
You can find more about my work below, on my CV, or via LinkedIn, Twitter, ResearchGate, or GoogleScholar. You can contact me via n.c.kobis[at]gmail.com.
A few recent papers on AI and ethical behavior
The promise and perils of using artificial intelligence to fight corruption
Nature Machine Intelligence
with Christopher Starke, & Iyad Rahwan
In this paper (open access link; download), we examine the excitement around using artificial intelligence to tackle corruption. We argue that whether the hope is warranted or not depends largely on whether AI is used top-down (by governments) or bottom-up (e.g., by civil society actors). Top-down use poses risks such as the consolidation of existing political power structures, big brother effects, and suppression of political opposition.
Bottom-up potential lies in AI helping to keep governments in check and innovative approaches to motivate citizens for accountability efforts.
Bad Machines Corrupt Good Morals
Nature Human Behaviour
with Jean-Francois Bonnefon & Iyad Rahwan
In this paper (open access link), we outline how artificial intelligence (AI) agents can negatively influence human ethical behaviour. They discuss how this capacity of AI agents can cause problems in the future and put forward a research agenda to gain behavioural insights for better AI oversight.
In this op-ed for the LA Times, we discuss the paper.
The Psychology of Deepfakes
with Christopher Starke & Ivan Soraperra
This Psychology Today blog post accompanies our paper on deepfake detection. A previous blog post in Psychology Today argued that raising awareness about deepfakes might help people to detect them. Our experiment suggests that awareness alone is not enough.
Inhalte von Youtube werden aufgrund deiner aktuellen Cookie-Einstellungen nicht angezeigt. Klicke auf “Zustimmen & anzeigen”, um zuzustimmen, dass die erforderlichen Daten an Youtube weitergeleitet werden, und den Inhalt anzusehen. Mehr dazu erfährst du in unserer Datenschutz. Du kannst deine Zustimmung jederzeit widerrufen. Gehe dazu einfach in deine eigenen Cookie-Einstellungen.
Artificial intelligence versus Maya Angelou: Experimental evidence that people cannot differentiate AI-generated from human-written poetry
Computers in Human Behavior
with Luca Mossink
New natural language generation (NLG) algorithms, like GPT-2, allegedly generate human-like text across diverse domains. We conducted two pre-registered experiments (total N = 830) on human and machine behavior in the creative writing domain. Outputs by GPT-2 were either selected or randomly paired with human poems by novices (Study 1) or professionals (Study 2). In an incentivized version of the Turing Test, participants failed to reliably detect selected algorithmic creative text. People are averse to AI-generated poetry, independent of receiving information about the origin of the poem, or not.
Fooled Twice – People Cannot Detect Deepfakes But Think They Can
with Barbora Doležalová and Ivan Soraperra
Hyper-realistic manipulation of audio-visual content, ie, deepfakes, presents a new challenge for establishing the veracity of online content. In a pre-registered behavioral experiment (N= 210), we show that (a) people cannot reliably detect deepfakes, and (b) neither raising awareness nor introducing financial incentives improves their detection accuracy. We also find that, (c) people are biased towards mistaking deepfakes as authentic videos (rather than vice versa) and (d) overestimate their own detection abilities
List of Academic Publications
The promise and perils of using artificial intelligence to fight corruption
Nature Machine Intelligence
with Christopher Starke & Iyad Rahwan
openly available here
Bad Machines Corrupt Good Morals
Nature Human Behaviour
with Jean-Francois Bonnefon & Iyad Rahwan
openly available here
Collaborative dishonesty: A meta-analytic review
Psychological Bulletin
with Margarita Leib, Ivan Soraperra, Ori Weisel & Shaul Shalvi
openly available here
Fooled Twice – People Cannot Detect Deepfakes But Think They Can
iScience
with Barbora Doležalová & Ivan Soraperra
Artificial intelligence versus Maya Angelou: Experimental evidence that people cannot differentiate AI-generated from human-written poetry
Computers in Human Behavior
with Luca Mossink
available here
The Consequences of Participating in the Sharing Economy: A Transparency-Based Sharing Framework
Journal of Management
with Shaul Shalvi and Ivan Soraperra
available here
Precision in Context Theory: In a Seller’s Market, Precise Prices Are Suboptimal.
Management Science
with Margarita Leib, Marc Francke, Shaul Shalvi, & Marieke Roskes
available here
Moral Uncanny Valley: A Robot’s Appearance Moderates How its Decisions are Judged. International Journal of Social Robotics
with Michael Laakasuo & Jussi Palomäki.
available here
Social norms of corruption in the field – Social nudges on posters can help to reduce bribery
Behavioural Public Policy
with Marleen Troost, Cyril Brandt, & Ivan Soraperra
available here
Recent approaches to the study of social norms and corruption
Book chapter: in A Research Agenda for Studies of Corruption
with David Jackson and Daniel Iragorri-Carter
available here
Intuitive Honesty Versus Dishonesty: Meta-Analytic Evidence.
Perspectives on Psychological Science
with Bruno Verschuere, Yoella Bereby-Meyer, David Rand, and Shaul Shalvi
available here
Anti-corruption through a social norms lens.
U4 Policy Issue
with David Jackson,
available here
Taxing the brain to uncover lying? Meta-analyzing the effect of imposing cognitive load on the reaction-time costs of lying.
Journal of Applied Research in Memory and Cognition.
with Bruno Verschuere, Yoella Bereby-Meyer, David Rand and Shaul Shalvi
available here
The road to bribery and corruption: Slippery slope or steep cliff?
Psychological Science
with Jan-Willem Van Prooijen, Francesca Righetti and Paul Van Lange
available here
The Social Psychology of Corruption.
Dissertation at VU Amsterdam
available here
Prospection in individual and interpersonal corruption dilemmas.
Review of General Psychology
with Jan-Willem Van Prooijen, Francesca Righetti and Paul Van Lange
available here
A Social Psychological View on the Social Norms of Corruption.
Book Chapter in: Corruption and Norms -Why Informal Rules matter
with Daniel Iragorri-Carter and Chris Starke
available here
Ein Atlas zur Unterscheidung von Korruptionsformen.
Book chapter in: Korruptionsbekämpfung vermitteln Didaktische, ethische und inhaltliche Aspekte in Lehre, Unterricht und Weiterbildung
with Oksana Huss
available here
“Who doesn’t?”—The impact of descriptive norms on corruption.
PLOS One
with Jan-Willem Van Prooijen, Francesca Righetti and Paul Van Lange
available here
Money for microbes—Pathogen avoidance and out‐group helping behaviour.
International Journal of Psychology
with Michael Laakasuo, Jussi Palomäki, and Markus Jokela
available here
False consensus in situational judgment tests: What would others do?
Journal of Research in Personality
with Janneke Oostrom, Richard Ronay, and Michael Cremers
available here
Why did the Panama Papers (not) shatter the world? The relationship between Journalism and Corruption.
Conference Proceedings How to research corruption? Conference Proceedings: Interdisciplinary Corruption Research Forum
with Chris Starke
available here
Intercultural Studies
Affective interpersonal touch in close relationships: a cross-cultural perspective
Personality and Social Psychology Bulletin
available here
Sex Differences in Mate Preferences Across 45 Countries: A Large-Scale Replication.
Psychological Science
with Katy Walter, Daniel Conroy-Beam, David Buss, Kelly Asao, Agnieszka Sorokowska, et al.
available here
Sex differences in human mate preferences vary across sex ratios
Proceedings of the Royal Society B
with Katy Walter, Daniel Conroy-Beam, David Buss, Kelly Asao, Agnieszka Sorokowska, et al.
available here
Contrasting Computational Models of Mate Preference Integration Across 45 Countries.
Scientific Reports
with Daniel Conroy-Beam, David Buss, Kelly, Agnieszka Sorokowska, et al.
available here
Assortative mating and the evolution of desirability covariation.
Evolution and Human Behavior
with Daniel Conroy-Beam, James Roney, Aaron Lukaszewski, David Buss, et al.
available here
Cross-cultural differences in a global “survey of world views".
Journal of Cross-Cultural Psychology
with Gerard Saucier , Kenner, J., Metaferia, H. et al.
available here
Data
In the spirit of #openscience, data for the empirical studies are freely available on my Open Science Framework page
Consulting and Policy Advice
I have become a member of the UNODC Education for Justice Expert Round-table that aims to foster tertiary education on anti-corruption.
Transparency International has been kind enough to invite me to join Transparency International’s Anti-Corruption Solutions and Knowledge Network of Experts
I have been invited to present my work at the Danish Foreign Ministry, the OECD Global Anti-Corruption and Integrity Forum, the Inter-American Development Bank and the European Risk Forum
I have also co-authored policy reports for the Dutch Tax Authority and the Anti-Corruption Resource Center (U4).
Anti-corruption through a social norms lens.
U4 Policy Issue
with David Jackson,
available here
Together with Shaul Shalvi, Uri Gneezy and Theo Offerman I co-authored a report for the Dutch Tax Authority called Increasing Ethical Conduct: A Behavioral Ethics Approach.
Media
Co-host of: KickBack the Global Anti-Corruption Podcast
Interviews and media features
Das Spiel für eine bessere Welt
Zeit Online
with Cyril Brandt & Christopher Starke
Why do we Collude with Corruption?
BBC World Service
Cyber snooping your therapist; Performing anxiety; Insiders' Guide; Bribery and corruption
BBC Radio 4
Social Psychology Videos: Social Dilemmas
InMind
with Paul van Lange & Hadija Mkwella
Leaping into Corruption
Scientific American
with Paul van Lange & Daniel Yudkin
Die Psychologie der Korruption, Radiofeature
Bayerischer Rundfunk
with Daniela Remus
Sometimes Corruption happens all at once
NY Mag
Does Corruption Happen Slowly, or All at Once?
The Atlantic
Oh Good: People Are More Likely to 'Abruptly' Become Corrupt, Study Finds
Vice Media
We Don’t Gradually Glide Into Corrupt Behavior—We Jump Head First
Pacific Standard
Corruptie heeft vaak te maken met plotse opportuniteiten
Express Business
Veja Corrupção surge de uma vez e não de maneira gradual, diz estudo
Hoje Mais
Cientistas descobrem que corrupção não surge aos poucos, mas de uma vez
Folha de S. Paulo
Social Psychology of Corruption
What are the social psychological forces of corruption?
How do others shape what we think is right and wrong?
Can we overcome societal challenges such as corruption by using behavioral insights?
These are some of the questions that I have spent the last 8+ years thinking, exchanging, writing, talking and researching about. Below you can see what I have come up with so far...
Some interim summaries
In 2012, a friend, Michael Laakasuo and I were discussing research ideas over a coffee.
Inspired by game theory and interdependence theory, we started modelling games for Paul
Van Lange’s expert workshop. We tweaked some aspects here, changed some parameters
there. After a second or third coffee, we re-examined what we came up with.
“This is corruption”
One of us said, looking at the game that I had drawn up. That is how corruption
came into my (academic) life. The fascination for the topic has not faded but has rather grown. After this initial coffee talk, I kept encountering corruption over and over again: in the
news, during conversations with friends, colleagues and taxi drivers or when reading
ostensibly unrelated literature.
So after more than five years of researching corruption and after having written my dissertation on the social psychology of corruption, I guess it is time to draw some interim conclusions and dare to spell out some of the main insights I have gained so far.
Corruption is an umbrella term that requires specification.
The word corruption is (often loosely) used to describe a wide array of phenomena. Academics, journalists and the public parlour use corruption to describe various behaviors ranging from bribery to embezzlement, from lobbyism to nepotism, from match-fixing to any type of misbehavior by political decision-makers. Researching, understanding and eventually curbing corruption requires closer differentiation. In my work I have introduced a distinction between individual and interpersonal form of corruption. From a psychological perspective it makes a big difference whether I abuse power alone (think of embezzlement) or in collaboration with others (like bribery). Together with Oksana Huss, I have worked on an Atlas of Corruption Types, for which you can find an illustration here.
The social element of corruption bears immense importance
My empirical work has extensively studied the social element of corruption and other forms of unethical behavior. My work on social norms of corruption examines the perceptions that people have about their social environment. These perceptions crucially shape the decision to engage in corruption. Social norms may also help to explain the vast differences in
corruption levels around the globe and provide a new lens for anti corruption.
A recent meta-analysis that I conducted together with Shaul Shalvi, Bruno Verschuere, Yoella Bereby Meyer and David Rand shows that people are less intuitively inclined to break unethical rules when it directly harms others.
Corruption over time: steep cliff or slippery slope?
One question that has always fascinated me is how people end up engaging in behaviors that they deemed unimaginable in the past. In particular the work of Albert Bandura and Harald Welzer inspired me to put the slippery slope idea to a test when it comes to corruption. In these studies we find that severe forms of corruption rather emerge abruptly (“steep-cliff-effect”) than gradually (“slippery-slope-effect”). Here, people rather seem to leap into corruption than do so gradually. Although providing some first answers, some fascinating the questions remain to be answered: how does power corrupt over time? How do people gradually normalize corrupt practices? Pursuing answers to these questions still fascinates me.
Interdisciplinarity helps to understand corruption
interdisciplinary research exchange is invaluable to gain insights into the complex dynamics of corrupt behavior. Fortunate enough to meet amazing corruption researchers from other disciplines, we together founded the Interdisciplinary Corruption Research Network. Starting at the Volkshotel in Amsterdam in 2015, so far we have organized conferences dedicated to corruption research across disciplines on an annual basis.
