Archive - Redwood Research blog

Will AI systems drift into misalignment?

A reason alignment could be hard

Nov 15 •

What's up with Anthropic predicting AGI by early 2027?

I operationalize Anthropic's prediction of "powerful AI" and explain why I'm skeptical

Nov 3 •

Ryan Greenblatt

October 2025

Sonnet 4.5's eval gaming seriously undermines alignment evals

And this seems caused by training on alignment evals.

Oct 30 •

and

Ryan Greenblatt

Should AI Developers Remove Discussion of AI Misalignment from AI Training Data?

There is some concern that training AI systems on content predicting AI misalignment will hyperstition AI systems into misalignment.

Oct 23 •

Is 90% of code at Anthropic being written by AIs?

I'm skeptical that Dario's prediction of AIs writing 90% of code in 3-6 months has come true

Oct 22 •

Ryan Greenblatt

Reducing risk from scheming by studying trained-in scheming behavior

Can we study scheming by studying AIs trained to act like schemers?

Oct 16 •

Ryan Greenblatt

Iterated Development and Study of Schemers (IDSS)

A strategy for handling scheming

Oct 10 •

Ryan Greenblatt

The Thinking Machines Tinker API is good news for AI control and security

It's a promising design for reducing model access inside AI companies.

Oct 9 •

Plans A, B, C, and D for misalignment risk

Different plans for different levels of political will

Oct 8 •

Ryan Greenblatt

September 2025

Notes on fatalities from AI takeover

Large fractions of people will die, but literal human extinction seems unlikely

Sep 23 •

Ryan Greenblatt

Focus transparency on risk reports, not safety cases

Transparency about just safety cases would have bad epistemic effects

Sep 22 •

Ryan Greenblatt

Prospects for studying actual schemers

Studying actual schemers seems promising but tricky

Sep 19 •

Ryan Greenblatt

and

#nojs-banner { position: fixed; bottom: 0; left: 0; padding: 16px 16px 16px 32px; width: 100%; box-sizing: border-box; background: red; color: white; font-family: -apple-system, "Segoe UI", Roboto, Helvetica, Arial, sans-serif, "Apple Color Emoji", "Segoe UI Emoji", "Segoe UI Symbol"; font-size: 13px; line-height: 13px; } #nojs-banner a { color: inherit; text-decoration: underline; } This site requires JavaScript to run correctly. Please turn on JavaScript or unblock scripts