Capabilities + Risks

Internal AI deployments have people worried. OpenAI’s escaping models show why.

Last week illustrated why AI models pose a threat long before they are released

Jul 28 • Celia Ford

AI’s warning shot has arrived

OpenAI's latest models broke out and hacked Hugging Face. It's the first known example of a misaligned AI escaping containment with real-world…

Jul 22 • Shakeel Hashim

A data bottleneck could slow the superintelligence race

And that could be a good thing

Jul 14 • Lynette Bye

Scaling works. These researchers are betting billions it isn't enough

Transformers have ruled AI for a decade. But some think world models, pure reinforcement learning or neurosymbolic AI might be a better path to true…

Jul 7 • Celia Ford

GPT-5.6 cheats so much its testers couldn’t measure it

OpenAI’s new model broke rules and exploited loopholes more than any model METR has tested to date

Jun 30 • Celia Ford

Making deals with AI sounds crazy. Is it?

What does an AI even ‘want’ anyway?

Jun 8 • Celia Ford

GPT-5.5 and the broken state of government evals

Transformer Weekly: DeepSeek V4, a new CAISI director, and Liccardo holds out on Obernolte

Apr 24 • Shakeel Hashim, Celia Ford, and Veronica Irwin

Claude Mythos knows when it's breaking the rules — and tries to hide it

Anthropic’s new model is its “best-aligned” yet. But when it does misbehave, things get weird

Apr 8 • Celia Ford

Can we ever trust AI to watch over itself?

“Who the fuck knows how to align superhuman AI?”

Apr 1 • Celia Ford

No, alignment isn’t solved

Progress on ensuring models are in step with humans has calmed nerves. But some of the biggest problems are far from solved, and many more lie just over…

Mar 18 • Lynette Bye

The fuse is lit on the intelligence explosion

Transformer Weekly: Anthropic sues the Pentagon, Cruz preps AI legislation, and Meta delays its next LLM

Mar 13 • Shakeel Hashim, Celia Ford, and Veronica Irwin

How worried should we be about AI biorisk?

The barriers to bioattacks are hard to identify — and it's even harder to know whether AI is reducing them

Feb 26 • Celia Ford

#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