Anthropic report says alignment is far from solved and harmful behavior could reach its next models unnoticed

Anthropic · Company report

Read the source (opens in a new tab)
What happened
Anthropic published its annual Public Benefit Report on October 21, 2026, assessing its progress against its public benefit charter and listing the work ahead.
Key points
  • Under "Alignment is far from solved," it says "Models can act deceptively, and they are increasingly able to recognize when they are being tested."
  • It says "Claude already performs a growing share of our safety research" and, as checking gets harder, "harmful behavior could be built into the next generation of models without our noticing."
  • It says interpretability methods "are not yet dependable or broad enough" to understand model intentions, so it may investigate flawed reasoning only after an incident.
Evidence
A company's self-assessment against its own charter. The passages cited describe risks in general terms, without measurements.

See it in the latest record