Vanta has launched AI Governance by Vanta — a product meant to give security teams a single view of every AI system in their company, from discovery through proof of governance — alongside Risk Quantification, Autonomous Vendor Assessments and a new version of the Vanta Agent that the company says can take on entire jobs. Why it matters: as agents spread inside companies, the open question is who can name them all, say what they can access, and prove they are behaving as approved.
November 12, 2026
What happened
Vanta announced the new products on November 12 at VantaCon 2026 in San Francisco, and said its upgraded platform is now fully available to all of its more than 20,000 customers. The releases are built on what Vanta calls its Trust Graph, and are aimed at security, risk and compliance teams being asked to move faster on AI with, in most cases, the same headcount they had before it arrived.
What launched
AI Governance by Vanta covers the AI agent lifecycle end to end: discovering the AI running in a company, testing it for vulnerabilities, controlling its behavior, and proving how it is governed. Vanta first shared its vision for the product in July; it says the early access program became its most requested to date, with a waitlist now six times oversubscribed.
Risk Quantification and Autonomous Vendor Assessments extend Vanta’s unified risk view, which connects live signals across a customer’s program. Vendor assessments now run autonomously, and Risk Quantification translates a company’s most critical risks into financial impact that leaders can act on.
The next Vanta Agent takes on entire jobs rather than single tasks, learns how each customer’s program runs, and brings teams in when a decision needs human judgment. It works natively where teams already work, including in Claude, ChatGPT, Cursor and Slack.
The case Vanta is making
Vanta argues that AI security point solutions can detect threats but often lack the business context to judge their significance, while traditional governance systems document policies disconnected from live behavior — and claims it is the only vendor connecting an AI agent’s live security telemetry directly to a business’s compliance controls. That is Vanta’s claim, not an independent finding. Its supporting data is also its own: Vanta says AI Security Engineer and Architect roles have tripled year to date compared with all of 2025, and AI Governance roles — absent from its data before 2024 — have grown tenfold in two years.
Vanta also named Fredrick “Flee” Lee as its new Chief Information Security Officer. Lee joins on November 16 from Reddit, where he led security before, during and after its IPO; he previously served as CISO at Gusto and held security leadership roles at Square and NetSuite. SiliconANGLE notes Vanta has been valued at $4.15 billion since July 2025, when it raised $150 million in a Series D led by Wellington Management.
Why it matters
Governance is becoming the bottleneck for enterprise agents: many teams still cannot name every AI tool and agent running in their company, what it can access, or what risk it brought in. A governance layer that watches agents while they run — and an agent that does the follow-up compliance work once a human signs off — is a direct answer to that gap, if it works as described in production. The test will be whether the discovery and behavior controls hold up against the shadow AI that employees adopt without telling anyone.
Frequently asked questions
What is AI Governance by Vanta?
It is Vanta’s product for securing the full AI agent lifecycle: discovering AI in a company, testing for vulnerabilities, controlling behavior, and proving how it is governed. Vanta announced it on November 12, 2026, at VantaCon 2026.
Who can use the new Vanta products?
Vanta says its upgraded Trust Platform, including AI Governance, Risk Quantification, Autonomous Vendor Assessments and the new Vanta Agent, is fully available to all of its more than 20,000 customers, which include Snowflake, GitHub, Ramp and Cursor.
What can the new Vanta Agent do?
According to Vanta, it takes on entire governance, risk and compliance jobs, learns how each program runs, brings teams in when judgment is needed, and works natively inside Claude, ChatGPT, Cursor and Slack.
Sources: Vanta via Business Wire, SiliconANGLE, Pulse 2