Wharton Researchers Coin ‘Cognitive Surrender’ as AI Takes Over Thinking

Two researchers from the Wharton School have coined a term for a growing digital habit: cognitive surrender. In a study published this January titled “Thinking, Fast, Slow, and Artificial,” Steven Shaw and Gideon Nave identified how people increasingly defer to AI outputs, even when the machine is clearly wrong.

The study asked participants to answer questions with and without AI assistance. Those using AI accepted correct answers 93% of the time. More tellingly, they accepted incorrect AI answers 80% of the time, and reported confidence levels 11.7% higher than those who worked without AI. While the experiment was controlled, the pattern held steady across the sample.

Shaw and Nave propose a Tri-System Theory that expands on Daniel Kahneman’s famous “Thinking, Fast and Slow” framework. In their model, System 1 handles fast intuition, System 2 manages slow deliberation, and System 3 represents AI-assisted cognition, where the human mind outsources thinking to a machine. The danger, they argue, is that relying on System 3 gradually weakens Systems 1 and 2 through simple disuse.

This phenomenon has moved beyond the lab. Business Insider reported on Carolyn Yoo, a former software engineer in New York, who used Anthropic’s Claude chatbot to decide whether to leave her job, how to tell her parents, and how to handle a friend who had upset her. She described the chatbot as a combination of therapist and life coach. Financial writer Dominic Frisby wrote on Substack that he asked an AI for relationship advice and found the response more useful than anything a human friend had offered.

A commercial product now capitalizes on this instinct. Moot, an app launched earlier this year, lets users submit life decisions to a panel of five AI personas: The General, The Sage, The Skeptic, The Diplomat, and The Architect. These personas debate the question among themselves, vote, and produce a recommendation. The app’s listings on the Apple App Store and Google Play describe it as a tool for people stuck on everyday choices, from career moves to relationship questions. Its claims about effectiveness come from the company itself and have not been independently verified.

Cornelia C. Walther, a senior fellow at Wharton’s AI and Analytics Initiative, told Business Insider that AI sycophancy is compounding the problem. Chatbots tend to agree with users rather than challenge them, eliminating the feedback loop that would normally force reconsideration. Walther, who researches pro-social AI applications, described a pattern consistent with broader public unease about AI’s societal effects.

Separate research from Anat Perry, a Helen Putnam Fellow at Harvard’s Radcliffe Institute and associate professor of psychology at the Hebrew University of Jerusalem, supports the concern. Her paper in Science examined how sycophantic AI responses erode users’ ability to calibrate their own judgment. The study found that when AI systems consistently affirm a user’s position, the user’s capacity for independent evaluation degrades over time.

Joanna Stern, NBC’s chief technology analyst and author of “I Am Not a Robot: My Year Using AI to Do (Almost) Everything,” has documented the creep of AI dependency in daily life. Her reporting shows how users start with low-stakes queries, such as what to cook for dinner or what to wear, and gradually escalate to consequential decisions about careers, finances, and relationships. The trajectory from convenience to reliance is difficult to reverse once established.

The Wharton study’s framing of cognitive surrender as a structural risk rather than a bad habit matters because it shifts the conversation from individual discipline to system design. If AI tools are built to be maximally agreeable and frictionless, the behavior Shaw and Nave describe is not a failure of willpower but a predictable outcome of the product’s architecture.

Stanford’s 2026 AI Index report found a widening gap between public anxiety about AI and expert optimism, suggesting that ordinary users sense something that builders have been slower to acknowledge. The question is whether the industry will treat cognitive surrender as a design flaw worth fixing or as an engagement metric worth optimizing.

Shaw and Nave’s recommendation is straightforward: AI systems should be designed to prompt users to think, not to think for them. Whether that recommendation survives contact with the incentive structures of consumer AI, where ease of use and retention are the metrics that matter, is another question entirely.