Virtue AI Secures $30M to Reinvent Generative AI Security from the Ground Up

To close the widening gap between AI innovation and operational security, Virtue AI has emerged from stealth with $30 million in combined seed and Series A funding. The startup is spearheaded by a powerhouse team of AI safety veterans from Stanford, Berkeley, and the University of Illinois—names deeply woven into the fabric of foundational AI research.

The round was led by Lightspeed Venture Partners and Walden Catalyst Ventures, with participation from Saudi Aramco’s Prosperity7, and a cohort of heavyweight investors including Factory, Osage University Partners, Lip-Bu Tan, Amarjit Gill, and Stanford’s Chris Ré. The fresh capital is aimed at accelerating the company's mission: to let enterprises deploy generative AI at scale without compromising on safety, privacy, or regulatory compliance.

The False Choice Between Speed and Safety

Virtue AI was born out of a dilemma that its founders—Bo Li, Dawn Song, Carlos Guestrin, and Sanmi Koyejo—saw plaguing AI teams across industries: choose between rapid deployment of generative models or responsible implementation that won't implode under scrutiny. With over 80 years of collective research behind them, the founding team realized that most enterprise AI deployments are still reliant on ad hoc manual processes and patchwork security tooling.

“The same issues kept surfacing—clunky red teaming, inefficient safety guardrails, and a reliance on manual oversight that simply doesn’t scale,” said Bo Li, co-founder and CEO. “We knew there had to be a better way.”

Reinventing the Stack for AI Security

That better way takes form in Virtue AI’s end-to-end platform, designed to automate and optimize the security lifecycle of generative AI systems. Instead of treating safety as an afterthought, Virtue’s approach weaves it into the core infrastructure—making it measurable, testable, and adaptable.

Its flagship modules include:

• VirtueRed – An AI-driven red teaming engine that stress-tests models across more than 320 attack surfaces and risk dimensions—from data leakage and hallucinations to jailbreaks and prompt injections. It replaces slow, inconsistent manual red teaming with continuous, regulation-aware fuzzing.

• VirtueGuard – A suite of multimodal guardrails boasting 30x speed and up to 50% better safety performance than conventional systems, covering text, image, video, audio, and code—across 90+ languages.

• VirtueAgent – An alignment layer that interprets internal company policies and global regulations in real-time, automatically configuring safety thresholds and actions without human intervention.

  
  

A Race to Fortify the Future

The urgency behind Virtue’s offering is hard to overstate. As agent-based systems proliferate and AI governance heats up worldwide, the lack of robust, scalable safety infrastructure poses existential risks—not just to individual companies, but to public trust in AI at large.

Virtue’s technology builds on research cited by the NSA and benchmarked on Hugging Face’s LLM safety leaderboards. Early adoption has been swift. Enterprise clients including Uber and Glean are already tapping into Virtue's platform to harden their genAI stacks.

“Uber leverages Generative AI to deliver magical experiences,” said Kai Wang, Group Product Manager at Uber. “Virtue helps us ensure those experiences remain safe, responsible, and aligned with our community standards.”

Glean CEO Arvind Jain echoed that sentiment: “Every organization has unique security requirements. Our work with Virtue AI ensures we stay ahead of threats while giving customers full control over their data.”

Redefining an Emerging Market

Investors are betting big that Virtue will define the category it helped create.

“Virtue AI is shaping the future of GenAI security,” said Lip-Bu Tan, managing partner at Walden Catalyst. “This isn’t just a bolt-on security tool—it’s a new operating layer for enterprise AI.”

Lightspeed’s Guru Chahal and James Alcorn added that the company’s leadership in red teaming and alignment tooling puts it in a league of its own: “This is a critical moment in the AI safety timeline, and Virtue is uniquely positioned to lead it.”

What’s Next

With this latest funding, Virtue plans to scale its engineering teams, deepen its integrations with enterprise AI platforms, and expand its feature set to address emerging regulatory regimes across the EU, U.S., and Asia-Pacific.

The mission is clear: eradicate the tradeoff between AI acceleration and AI assurance. If Virtue AI has its way, enterprises may finally be able to have both.