Virtue AI is an AI security company founded by leading academics Bo Li, Dawn Song, Carlos Guestrin, and Sanmi Koyejo to resolve the tension between rapid AI innovation and enterprise-grade safety. Its unified platform comprises three core modules: VirtueRed, an algorithmic red-teaming engine that checks more than 320 risk categories; VirtueGuard, a suite of guardrail models said to run 30× faster while delivering 40-50 % better performance across modalities and 90+ languages; and VirtueAgent, which automatically interprets internal and regulatory policies to enforce security controls. The technology stems from award-winning research recognized by the NSA and is already being adopted by customers such as Uber and Glean. By addressing threats like hallucinations, prompt injection, privacy leakage, and data poisoning, the company targets the rapidly expanding, multi-billion-dollar AI security market. Fresh capital will be used to extend product capabilities, scale go-to-market efforts, and recruit additional AI-safety talent. Virtue AI is headquartered in San Francisco and is still pre-revenue disclosures, but early customer traction underscores growing enterprise demand. To date it has raised $30 million in external funding.

Total raised

$30M

Funding rounds

1

Latest round

Seed

Latest activity

Apr 2025