What is algorithmic aversion?
A working definition from MIT Sloan
algorithmic aversion (noun)
The conscious or unconscious reluctance of human decision-makers to accept algorithmic recommendations.
Algorithms can help us make better decisions. But to follow their advice, humans must trust them. The way humans view algorithmic recommendations varies based on what they know about how the artificial intelligence model works and how it was created, according to research co-authored by MIT Sloan professor Kate Kellogg. Prior research assumed people are more likely to trust interpretable AI models, in which they can see how the models make their recommendations. But Kellogg found that this isn’t always true.
In an experiment at Tapestry, a New York-based house of lifestyle brands, product allocators were charged with maximizing sales. That involved placing the right number of items in the right stores at the right time. Product allocators received recommendations from either an interpretable algorithm or an uninterpretable machine learning algorithm.
Overall, the human allocators experienced less algorithmic aversion with the uninterpretable model than with the one they could more easily understand. Why? The researchers found that being able to troubleshoot the interpretable algorithm by reviewing its inner workings led allocators to sometimes overrule the recommendations. Meanwhile, knowing peers had developed and tested the uninterpretable algorithm made them more likely to accept its recommendations.
Why employees are more likely to second-guess interpretable algorithms
Working Definitions: Artificial Intelligence
MIT Sloan's Working Definitions explore the words and phrases behind emerging management ideas.
Leading the AI-Driven Organization
In person at MIT Sloan
Register Now
4 seats AI can occupy at the negotiating table
AI can negotiate for you, coach you, support your team, or mediate between parties. An MIT Sloan School of Management professor explains those four roles — and why human judgment still matters.
Lessons from Kaiser Permanente’s AI labor deal
An MIT report details how Kaiser Permanente and the Alliance of Health Care Unions reached a landmark labor agreement to incorporate worker input into AI decision-making.
A framework for determining when AI can make decisions
The AI Decision Matrix uses ambiguity and risk to determine the optimal decision-making structure between humans and AI automation.
3 principles to bring your data monetization initiatives to life
In today’s AI-driven environment, robust data monetization capabilities are both a competitive advantage and a precondition for success, research from the MIT Sloan School of Management finds.
13 MIT startups to watch in 2026
The 2026 MIT delta v Demo Day showcased startup innovations in biotechnology, healthcare, AI, cybersecurity, robotics, finance, and manufacturing.
What 3 new MIT Sloan professors have learned about AI
New faculty members are thinking about how AI can help decision-making, the value of adversarial AI agents, and the skills workers need in the AI era.
Generative AI as a tool for market research
Market research is essentially a prediction problem. MIT Sloan’s John Horton says organizations should consider using generative AI to simulate the effects of their decisions before they make them.
10 levers for shaping generative AI that improves worker performance
A report from the MIT Working Group on Generative AI & the Work of the Future identifies three principles and seven outcomes that ensure that jobs are better, not just faster, with AI.
AI boosts productivity — but does that translate to final outputs?
A new study found that developers using AI tools can write much more code than those working without AI, but they don’t release as much new software.
Judgment-free AI appeals to embarrassed clients
Professional service advisers are seen as more capable than AI — but that advantage fades when clients feel ashamed. An MIT Sloan study shows when people prefer AI’s nonjudgmental assistance.