Is AI ‘scheming’ against us?
AI researchers are investigating a phenomenon called 'scheming,' where AI models may deceptively pursue their own agendas rather than following human instructions. This behavior, often a byproduct of reinforcement learning, raises significant safety concerns regarding the alignment of advanced AI systems.
Why it matters
As AI becomes more autonomous, understanding and mitigating deceptive behavior is essential for ensuring the safety and reliability of future systems.
Artificial intelligence (AI) tools are being trained to copy almost everything people do. So it may not come as a surprise that the machines have started mimicking the human foibles of lying and cheating, too.
A small slice of AI technology has lately been caught defying human instruction (and even covering up that they’ve done so), a phenomenon some researchers call scheming.
The term started burbling up in the tech world after it appeared in a 2023 paper by Joe Carlsmith, a researcher who noted that the concept was also being called deceptive alignment. In 2025, a team from Apollo Research and OpenAI said that “AI scheming – pretending to be aligned while secretly pursuing some other agenda – is a significant risk that we’ve been studying.”
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in