Exclusive: Meta poaches Virtue AI bigwigs to boost security
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Sarah Grillo/Axios. Stock: Getty Images
Meta Superintelligence Labs is hiring three founders and other team members behind AI security startup Virtue AI, Axios has learned.
Why it matters: Amid a swell of AI safety concerns following the release and pullback of Anthropic's latest models, Meta is looking to beef up its security efforts around AI agents.
Zoom in: Virtue AI co-founders Bo Li, Dawn Song and Sanmi Koyejo, leaders in AI security, are joining Meta Superintelligence Labs, according to an internal post viewed by Axios.
- The hires also include members of Virtue AI's broader team, which develops enterprise AI security tools spanning automated red teaming, runtime guardrails and AI governance.
- The terms of the arrangement were not disclosed.
What they're saying: "As we ship AI products to billions of people and build increasingly capable agents, keeping those systems safe, reliable, and trustworthy is foundational," the memo says.
- Li and Song will report into Nat Friedman, a Meta Superintelligence Lab executive.
- Sanmi Koyejo will report into Rob Fergus, head of FAIR, Meta's fundamental AI research lab.
Between the lines: AI labs are doubling down on agentic security amid increased scrutiny of which models the White House may pull and why.
- It's still unclear when users will get access to Anthropic's latest models, Fable 5 and Mythos.
- OpenAI expanded access to its cybersecurity tools this week and did not face similar scrutiny, Axios' Sam Sabin reports.
Zoom out: It's also the latest example of hiring a discrete team as opposed to an entire company.
- Meta poached the founders and team behind Dreamer, an AI startup focused on helping people build agents, in March.
The bottom line: Meta is expanding its security talent as the AI race heats up.
