Verify before you share

One-minute script

AI-generated language can sound polished even when a claim is unsupported or wrong. If a claim will influence a decision, a customer, or another document, fluency is not enough: check it against the best available original source.

Try this

Choose the claim that matters most and verify it before the material travels.

Ask together

Where would an error go next if we shared this?

Facilitator notes

Choose a claim participants can check without exposing confidential information. Distinguish an original source from another summary repeating the same assertion.

Source notes

The NIST Generative AI Profile identifies confident false content as a generative-AI risk and recommends evaluation, verification, and risk controls suited to the context.