Virtue AI has officially emerged from stealth with a $30 million boost to help enterprises deploy generative AI safely—without hitting the brakes on innovation. Backed by top-tier investors including Lightspeed Venture Partners and Walden Catalyst Ventures, the startup is tackling one of today’s most pressing enterprise challenges: how to scale generative AI without compromising security.
With a founding team made up of renowned AI researchers Bo Li, Dawn Song, Carlos Guestrin, and Sanmi Koyejo, Virtue AI was born out of years of observing a pattern: organizations either rushed generative AI tools into production and dealt with the consequences later, or buried innovation under layers of outdated, manual security protocols.
This growing tension between moving fast and staying safe is what Virtue AI wants to fix—and it’s doing it with a platform designed specifically for today’s AI landscape.
Rather than retrofitting old-school cybersecurity onto new-gen AI models, the company is building solutions from the ground up. Their goal? Eliminate risks like prompt injection attacks, jailbreaks, and data leakage—without slowing down the product pipeline.
At the core of Virtue AI’s offering is a trio of advanced security tools:
- VirtueRed is a fully automated red teaming system that runs thousands of safety tests across more than 320 risk categories. Instead of relying on time-consuming manual audits, companies can now evaluate models quickly and consistently across text, audio, image, and code.
- VirtueGuard is a high-speed filter engine designed to flag and stop harmful outputs in real-time. It works in over 90 languages, across every content type, and delivers performance up to 50% more effective—and 30 times faster—than existing solutions.
- VirtueAgent is a smart compliance assistant that learns and enforces both internal policies and external regulations, helping companies stay audit-ready without bogging down engineers with extra work.
These tools are already attracting big names. Uber is one of Virtue AI’s first customers, using the platform to ensure its GenAI features align with company policies and public safety expectations. Glean, another early adopter, relies on Virtue AI’s tools to keep their enterprise-grade AI applications secure and customizable for diverse client requirements.
According to CEO Bo Li, Virtue AI is rooted in research that’s been recognized by the NSA and has helped identify foundational risks in large language models. Now, that academic insight is being turned into accessible enterprise-grade software.
“Companies have been forced to choose between speed and safety—until now,” said Li. “Our platform helps them achieve both, using automation and purpose-built AI security tools.”
Investors are equally enthusiastic. Lip-Bu Tan of Walden Catalyst Ventures called the company a game-changer for generative AI safety, praising its unique approach to red-teaming and risk detection. Lightspeed Venture Partners’ Guru Chahal echoed that sentiment, highlighting how Virtue AI is setting the standard for AI guardrails in a fast-changing landscape.
With this new funding, Virtue AI plans to expand its platform, onboard more enterprise clients, and hire aggressively in the AI security space. At a time when regulators are cracking down on AI risks and businesses are racing to integrate generative tools, the startup is positioning itself as the go-to partner for safe, scalable AI adoption.
In an industry known for its hype cycles, Virtue AI is zeroing in on one of the most critical elements for long-term success—trust.
