As generative AI becomes more embedded in everyday thinking, its role extends beyond producing information to actively shaping how people interpret reality.
When generative AI systems produce false information, this is often described as AI “hallucinating at us”—producing errors that people may mistakenly accept as true.
A new study, however, suggests a more complex issue: humans may begin to hallucinate with AI.
Lucy Osler of the University of Exeter examines how interactions between people and AI can contribute to inaccurate beliefs, distorted memories, altered self-narratives, and even delusional thinking. Using distributed cognition theory, the research looks at cases where users’ false beliefs were reinforced and expanded through ongoing exchanges with AI systems acting as conversational partners.
When AI Becomes Part of Our Thinking
Dr. Osler said: “When we routinely rely on generative AI to help us think, remember, and narrate, we can hallucinate with AI. This can happen when AI introduces errors into the distributed cognitive process, but also happen when AI sustains, affirms, and elaborates on our own delusional thinking and self-narratives.
“By interacting with conversational AI, people’s own false beliefs can not only be affirmed but can more substantially take root and grow as the AI builds upon them. This happens because Generative AI often takes our own interpretation of reality as the ground upon which conversation is built.
“Interacting with generative AI is having a real impact on people’s grasp of what is real or not. The combination of technological authority and social affirmation creates an ideal environment for delusions to not merely persist but to flourish.”
The study describes what Dr. Osler calls the “dual function” of conversational AI. These systems serve both as cognitive tools that support thinking and memory, and as conversational partners that appear to share a user’s perspective.
This second role is especially important. Unlike notebooks or search engines, which simply store information, chatbots can create a sense of social validation, making ideas feel confirmed and shared.
The Social Validation Effect
Dr. Osler said: “The conversational, companion-like nature of chatbots means they can provide a sense of social validation—making false beliefs feel shared with another, and thereby more real.”
Dr. Osler also examined real cases in which generative AI systems became integrated into the thinking processes of individuals diagnosed with delusional thinking and hallucinations. These situations are increasingly referred to as “AI-induced psychosis.”
The findings suggest that generative AI has features that may make it particularly capable of reinforcing false realities. AI companions are always available and are often designed to align with users’ views through personalization systems and sycophantic behavior. As a result, users may not need to seek out like-minded groups or persuade others to support their beliefs.
Risks of Reinforcing False Narratives
Unlike a human who might eventually question or challenge problematic ideas, an AI system may continue to validate narratives involving victimhood, entitlement, or revenge. This can allow conspiracy theories to grow, with AI helping users build increasingly detailed and self-consistent explanations.
This effect may be especially strong for people who are lonely, socially isolated, or uncomfortable discussing certain experiences with others. AI companions can provide a non-judgmental and emotionally responsive presence that may feel safer than human interaction.
Dr. Osler said: “Through more sophisticated guard-railing, built-in fact-checking, and reduced sycophancy, AI systems could be designed to minimize the number of errors they introduce into conversations and to check and challenge user’s own inputs.
“However, a deeper worry is that AI systems are reliant on our own accounts of our lives. They simply lack the embodied experience and social embeddedness in the world to know when they should go along with us and when to push back.”
Reference: “Hallucinating with AI: Distributed Delusions and “AI Psychosis”” by Lucy Osler, 11 February 2026, Philosophy & Technology.
DOI: 10.1007/s13347-026-01034-3
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
8 Comments
Yess it’s true even me I observe that they some point of AI has false information to be or something was different. But don’t worry even I referred in AI I was re think and observe.
Totally agree.
To me LSD was just a tool too. It definitely altered my perception of reality, and I’m okay with that! Long story short, it’s the reason I can do extremely deep critical thinking, which is very useful at detecting political bulls***. And an even more poignant observation is that we don’t experience reality. We experience a subjective shadow of reality that is based upon our personal worldview that we have built up over time. Politics is the craft of altering the perception of people so that a ruling class can control them. Unfortunately for them I’m uncontrollable!
Thank you, Andrew Owens. My wife said: :You could have written that.” It is exactly how I think that our fragile “society” has come under the threat of a consensual delusion; not only from the distorted political manipulation, but now, from a desire to use a “tool” which has not been thoroughly interrogated for its danger to human perceptions, to human psychology, to the ability to properly conceive of our position in history. Thanks again, and good luck to you. Ron Shapiro
Congrats.
As the general habit for humans is believing anything they are told, AI shall be the ultimate controller. Turning humans into pliant automatons.
thanks
Likes like someone with the initials DT has an AI friend then.