SentinelOne's Up-to-$300M Prompt Security Deal Is the Only Priced Buy in a Trio of AI-Governance Acquisitions

"Up to $300M" is a ceiling. CB Insights does not say what SentinelOne paid at close for Prompt Security, and the two other governance acquisitions on its list, OpenAI's purchase of Promptfoo and Asana's of Stack AI, come with no price at all ^[2][3][4][9].

That leaves one capped number as the entire public price record for the category. Three deals across twelve months is one about every four months ^[12]. That cadence says the assets are moving. But the buyers do not repeat: three deals, three distinct acquirers, and none of them has come back for a second governance company ^[10].

The sell-side case is in Superblocks' own words. "AI made building fast, but building is 10% of the problem," said Puneeth Ghodgeri, the company's VP of marketing, who described the other 90 per cent as "shipping, securing, integrating, governing, deploying, and maintaining all of those apps at enterprise scale" ^[8]. Read from the buy side, that is a list of six operational jobs the enterprise customer is already paying an incumbent platform to do.

The three names CB Insights expects to take off next are Superblocks, Lyzr and Daytona. Superblocks is a 2026 AI 100 winner whose platform builds AI-built apps with governance, compliance and secure data connectors included ^[5]. Lyzr, at Series B, sells security and compliance controls around agents in production ^[6]. Daytona runs AI-generated code in isolated environments away from production systems ^[7]. The list carries no round size and no valuation for any of the three, and only Lyzr's stage is given ^[14]. A buyer would have to compute an entry price from something else.

So the evidence supports a target list with one price on it, and the two readings that follow are worth separating. If the unpriced deals were small, the category is being absorbed cheaply and early, and the $300M figure is the top of the range instead of the middle of it ^[9]. If they were large and undisclosed, the ceiling understates what governance tooling clears at, and a founder in the category is underpricing. Either reading changes when one of SentinelOne, OpenAI or Asana buys a second governance asset, and so far each of the three has bought once ^[10].

A single contingent maximum is not a comparable, because a comparable requires two prices paid, and the material has one price that may not have been paid in full ^[2][9].