Vanta Buys Pillar to Tie AI Governance to Actual-Time Controls

Vanta plans to buy a startup led by a former Aqua Safety product supervisor to attach synthetic intelligence governance with real-time safety enforcement.

See Additionally: Frictionless Safety: How AI is Redefining Entry in Extremely Regulated Industries

The San Francisco-based belief platform mentioned its proposed acquisition of Tel Aviv, Israel-based Pillar Safety will assist organizations uncover AI property, apply policy-based guardrails, block unsafe actions and produce audit-ready proof, mentioned GRC Senior Director Khush Kashyap. She mentioned the deal will be part of Vanta’s governance with Pillar’s agent discovery, monitoring, red-teaming and enforcement.

“Vanta helps hundreds of firms show their AI is reliable to the individuals shopping for from them,” Kashyap informed ISMG. “However firms working AI want the opposite half too, which is seeing what’s truly working and controlling it. That is what Pillar builds.”

Pillar, based in 2023, employs 33 individuals and closed a $9 million seed spherical in April 2025 led by Protect Capital. The corporate has been led since inception by Ziv Karliner, who beforehand spent greater than 4 years at Aqua Safety, together with two-and-a-half years as product supervisor. Calicalist reported that Vanta paid $150 million for Pillar, which an organization spokesperson disputed, although she declined to reveal the worth (see: Vanta Secures $150M at $4.15B Valuation to Advance AI Belief).

Imposing Boundaries on AI Brokers

Pillar’s aim from its inception was to guard organizations by the AI transformation, and Karliner mentioned working with Vanta provided a technique to develop adoption and protection shortly, whereas having access to Vanta’s safety and governance context. The acquisition will assist each firms advance their shared imaginative and prescient and serve a bigger buyer base, Karliner mentioned.

“The AI safety market is a really busy area,” Karliner informed ISMG. “There are loads of totally different angles to it, and as a younger firm, we’ve got to construct our fame on the analysis facet and with our prospects which are leaning closely into AI safety early on. The entire classes that we realized and the nice product that we received to construct collectively can now serve a a lot bigger buyer base.”

Vanta helps firms reveal when their AI is reliable, Kashyap mentioned, whereas Pillar addresses the complementary have to see what AI is doing and management it in actual time. Pillar’s capabilities are supposed to assist establish AI brokers accessing delicate knowledge or taking actions with out evaluation, which Kashyap mentioned are dangers that coverage paperwork alone can’t tackle.

“Pillar’s expertise lets prospects uncover AI brokers, crimson workforce them, monitor their habits and block unauthorized actions,” Kashyap mentioned. “If you mix that with Vanta’s visibility into property, customers, instruments, knowledge and management threat from the belief graph, that expertise will let CISOs block dangerous actions whereas permitting actions that solely look dangerous.”

AI brokers function at a tempo that makes quarterly or annual governance cycles inadequate, and Kashyap mentioned controls want to use at runtime. Vanta’s current position is to translate controls into proof, which Kashyap mentioned is accepted by auditors at greater than 20,000 firms. Pillar is anticipated to assist flip governance necessities into enforceable boundaries and block unauthorized actions in actual time.

“AI governance early entry is essentially the most requested program in Vanta’s historical past, six instances oversubscribed,” Kashyap mentioned. “And the every day Vanta agent customers are up by 253% within the final three quarters. Clients are asking for this now, and the market is prepared.”

Why Governance and Enforcement Are Higher Collectively

The central drawback is the hole between figuring out what insurance policies and controls require and seeing whether or not brokers are appearing in accordance with them, Kashyap mentioned. She summarized the hole as enforcement with out governance context producing alerts, whereas governance with out enforcement produces solely a paper path. The mixed method will create proof as occasions occur moderately than after the very fact.

“Enforcement with out governance context is like alerts,” Kashyap mentioned. “However governance with out enforcement is simply paper path. Typically individuals additionally name it compliance theater.”

Vanta prospects at the moment tackle AI governance by management design, insurance policies, documentation, threat assessments and frameworks, and Kashyap mentioned these approaches assist prospects perceive and doc their obligations, however are restricted in proactive, real-time enforcement. Shopping for Pillar will assist Vanta prospects transfer past receiving steerage on guardrails to imposing them, Kashyap mentioned.

“Now we’re governing the apps in seconds, so governance has to maneuver to runtime, and if we do not, we’re simply inflicting extra documentation and loads of paper path,” Kashyap mentioned.

Pillar’s discovery, red-teaming, monitoring, and blocking capabilities may very well be strengthened by Vanta’s visibility into property, customers, instruments, knowledge, dangers, controls, exceptions, insurance policies and checks, Kashyap mentioned. Clients can present a written AI coverage and use the mixed product to automate controls round it, which Karliner mentioned this might join coverage necessities to enforcement throughout coding brokers.

“GRC turning into enforcement is a name of the hour to be able to speed up all the things,” Karliner mentioned. “The entire safety applications are actually going to speed up and switch into runtime. Defending AI, enabling AI, enabling management and belief, is the best way to allow belief and to permit adoption of an accelerating expertise whereas minimizing threat, controlling the danger.”