Threat Pulse 62High
Anthropic and Theorem use Claude to formally verify 123 open-source libraries
Severity 1ResolvedconfirmedAI and software supply chainAI-related

Anthropic's Frontier Red Team reported machine-checked proofs (about 180 million lines of Rocq) that 239 fuzz-test entry points across 123 libraries cannot crash on inputs under 64 MB, finding and patching hundreds of bugs along the way. Anthropic argues that since AI can now find and exploit bugs cheaply, defenders need proofs rather than tests.

What to do: Proofs cover RISC-V static binaries only; Anthropic lists the limits.

AI angle: Defensive use of frontier models at scale.

· 14 days ago · 123 open-source libraries · Anthropic
SourcesAnthropicUnite.AI
On X
@AnthropicAIverifiedOct 22, 22:01 UTC

New on our Frontier Red Team blog: using Claude to formally verify software at large scale. In collaboration with @theoremlabs, we used Claude to prove security properties in over one hundred real-world open-source libraries over the span of a few weeks.

Open on X
Why this score

A defensive research result; included because it is the main counterweight offered to AI-driven offense.

Back to the board

Updated Nov 5, 01:46 UTC. Refreshes on its own.

Add a threat

Paste a link to an advisory, an article or an X post. A person reads every submission before it shows; nothing posts on its own.

Share

The picture people see when the link is posted
Post on XSave picture

This picture is what X, iMessage, Slack and Facebook show under the link. It is drawn fresh from the live board: the board and chart pictures update every 10 minutes, an event picture follows that event.