← All glossary entries

Authority Borrowing

Letting the model's confident tone carry a claim you have not verified.

How it feels

The claim feels safer because the sentence arrives with institutional calm.

At work

A report includes a confident statistic from AI because it sounds citation-ready, but no one finds the primary source.

Hidden cost

Decisions: confidence substitutes for verification.

You may have this if

Can you point to the primary source for the strongest claim in the output?

Try this instead

Use the primary-source rule: any claim that changes a decision must be verified outside the model.

Team rule

No AI-sourced claim enters a decision doc without a human-checked source link.

Primary-source verification and Decisions circuit.

Related

The site provenance page states that numeric claims in the manuscript were verified against primary sources.

Inference

Authority Borrowing names the opposite workflow: accepting confidence without source ownership.

Authority Borrowing: confident tone is not a source.

Open square card

If the claim matters, ask: where is the source? Not "does this sound right?" Source first. Confidence later.

AI hallucination confident toneverify ChatGPT claimsAI authority bias

Find where this pattern is hitting you first.

The Anti-AI Brain Score maps your weakest circuit, then points you back to the practices that reload it.

Take the 2-minute score → Use this in your team