AI supercharges formal verification: Anthropic proves 239 entry points never crash with 180M-line proofs
kfountou · x · 2026-11-18
A PL Perspectives essay analyzes why formal verification is having a moment, with AI as both problem and solution: agents find more bugs in security-critical software, while also completing verification tasks that once took humans years.
- Capital flooding in: Theorem outlined a $134M roadmap to verify nixpkgs; Harmonic and &Truth raised $295M and $70M respectively for automated verification.
- Landmark results: Anthropic and Theorem verified 239 entry points across 123 open-source libraries never crash using 180 million lines of machine-checked Rocq proofs; Google verified Signal's security properties in 6.5 million lines of Lean.
- The author, a former Facebook engineer whose internal verification effort stalled years ago, reflects on efficiency tradeoffs and the new role of formal methods researchers in an agent-driven world.
More from coding & agent
- Neon MCP Has Double the DAU of Its CLI Despite All Efforts Pushing CLI — JensHonack · 2026-11-18
- OpenAI's Agent-Swarm Playbook for Boring Spark Upgrades, Shown at QCon — AdiPolak · 2026-11-18
- Orthogonal Launches Data Marketplace for Agents With 100+ Per-Call Data Providers — ycombinator · 2026-11-18
- Replit CEO on Missions: an Agentic Swarm Harness That Hits Goals Within Budget — amasad · 2026-11-18
- User says he'd pay $20/month extra for an X API that lets agents read posts — willdepue · 2026-11-18
- HF CEO: RL-Adapted Qwen3.8-27B Doubles SWE-bench Pro Score to 62.7% for pi Agent — victormustar · 2026-11-18