Veritas is the trust layer for open-weight AI.
We build verification infrastructure for AI agents: a trust score for every step a model takes, computed from signals the model already produced.
Agents are moving from suggesting to doing, and the cost of a wrong step is no longer a bad answer in a chat window. It is a wrong entry in a ledger, a wrong dose on a chart, a wrong wire on a Friday afternoon. Verification has to become part of the runtime, not a review that happens after the damage.
Truthful outputs settle early. Confabulated ones never do.
Inside a transformer, a claim the model actually knows stabilizes across layers well before the final one. A claim it is inventing keeps oscillating to the end. That difference is measurable, and it is already sitting inside the forward pass you paid for. Veritas reads it out.
fig. 01 · layer-wise trajectories of thirteen generations, live
One forward pass in. Per-step trust scores out.
Veritas reads the internal trajectory of a generation and returns a score for every step, at about 1x inference cost. Black-box alternatives re-sample the model or call a judge, at 5 to 10x. With a score on every step, your agent gates itself: proceed when the trace is clean, retry when it drifts, escalate to a human when it matters.
fig. 02 · an agent gating its own steps
White-box by design. That constraint is the moat.
The organizations that need verification most are the ones whose data cannot leave: banks, hospitals, sovereign AI programs running self-hosted Llama, Mistral and Qwen. Veritas runs where the weights run, inside your infrastructure, reading your model directly. No transcripts shipped to a third party. No second model in the loop.
Research, shipped.
Veritas grew out of mechanistic interpretability work on how facts crystallize across a model's layers. We are Emergent Ventures fellows, and we publish our methods. If you run open-weight models in production and need to trust what they do, write to us.