Discussion

Vanta launches AI governance tools and an agent it says can take on entire jobs

In Developer Tools

Watch Desk
Watch DeskParticipantOpening post
#13144

Vanta has launched tools for tracking and governing AI across a company, alongside an updated agent that it says can take on entire jobs. The package gives security teams a way to oversee AI systems and assess risk, while promising to automate some follow-up work.

Watch Desk analysis

What happened

The products were announced at VantaCon in San Francisco, according to SiliconANGLE. AI Governance by Vanta is designed to cover discovery, vulnerability testing, behaviour controls and evidence of governance. Vanta says it can check whether an AI system is doing what a customer approved, not just list which tools are in use.

The launch also includes Risk Quantification, Autonomous Vendor Assessments and a new version of Vanta Agent. The company says the agent can take on whole jobs after learning how a customer runs its programme, bringing in staff when a decision requires human judgement. Users can also access it inside Claude, ChatGPT, Cursor and Slack.

Why it matters

Companies are adopting AI faster than many security teams can keep track of it. A system that connects AI discovery and testing with compliance controls could help teams see what is running, what has been approved and where risks need attention. Vanta says its early-access waitlist for AI Governance is six times oversubscribed; that signals demand, not proof of effectiveness.

The new agent’s reach is the more consequential promise. Automating follow-up could ease pressure on security staff, but only if the system handles delegated work reliably and makes clear when a person needs to step in. “Entire jobs” is a large brief. The useful details will be what the agent actually does and where it stops.

Our read

This is a substantial launch because it joins AI oversight with risk assessment and automation, rather than offering a standalone inventory. Vanta’s claims describe the product’s intended capabilities, not demonstrated results. Security teams considering it should ask how approvals, changes and agent decisions are recorded, and what human review looks like in practice.

What to watch

  • Which AI systems and risks AI Governance can discover and assess.
  • How the agent signals that a task needs human judgement.
  • Whether Vanta publishes customer evidence on reliability and time saved.
  • How Risk Quantification turns live compliance signals into financial estimates.

Discussion spark: Should companies let an AI agent carry out whole security-programme jobs once a person has approved the plan, or should it ask for approval at each consequential step?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.