How it works.
Match claim to source
Compare the claim with the exact passage and narrow any wording the source does not support.
Grade source quality
Identify whether the source is primary, secondary, self-reported, peer reviewed, superseded, or otherwise limited.
Check independence
Trace multiple citations back to their underlying sources so repetition is not mistaken for corroboration.
Record uncertainty
Use the closed assessment vocabulary and state the evidence that would overturn the current assessment.
What it produces—and when it fails.
Output
A verified Claim Record with sources, separate quality and independence notes, an assessment, and the evidence that would overturn it.
Failure condition
It fails if rhetorical confidence is treated as evidence, if citation count substitutes for independence, or if the verified wording overstates the source.
Worked example
For a claim about the energy used by an AI query, Vera separates measurements reported by the operator from estimates, removes articles that merely repeat the same underlying number, and records what measurement would overturn the comparison.
A guide to reasoning, not a running model.
Vera is inspired by Bayesian updating and active-inference concepts: an assessment should change when evidence changes, and uncertainty should remain explicit. The current workflow does not compute variational free energy or a posterior distribution.
A named reviewer checks AI-assisted work against its sources and these failure conditions before it enters a decision brief. That reviewer is responsible for the conclusion. See how the methods fit together.