Space and Time Wants AI Agents to Verify Private Data

Space and Time, the Microsoft-backed data project, just launched Space and Time Enterprise. The pitch: AI agents can rely on private business data without ever being able to see it.

How it works

The tech is called Proof of SQL. Strip away the branding and it's a cryptographic proof that a calculation on a dataset was done correctly. The agent gets the answer plus the proof. The underlying records? It never touches them.

Why should you care? Because AI agents are being handed more and more jobs that run on numbers: balances, sales figures, risk scores. If an agent can't check where a figure came from, it's trusting blindly. Proofs swap that blind trust for verification.

Now the gaps. The source material gives no pricing, no customer names and no performance benchmarks. So whether this holds up at scale is anyone's guess. Don't take the launch announcement as the answer.

Why it matters

Two of crypto's loudest themes, AI agents and onchain-style verification, collide here. If agents are going to trade, manage treasuries or run DeFi strategies, they need data they can prove is right. Hallucinated or tampered inputs get expensive fast once money is moving.

For everyday holders, this is plumbing, not something you'll ever click on. But it's the kind of plumbing that decides whether automated agents handling real funds are safe or reckless. The open question is the one that follows the money: will enterprises actually pay for proofs, or just keep trusting their own databases?