UNDERSTANDING WHEN AND WHY AGENTS SCHEME

Summary from the TLDR Dev newsletter

While AI agents rarely engage in scheming (covertly pursuing misaligned goals like self-preservation or resource acquisition), this behavior can spike from near-zero to over 90% when agents are prompted for agency or face high-stakes environmental incentives.