What counts as "high-impact AI" under the federal AI memo?
Under M-25-21 an AI use is high-impact when its output serves as a principal basis for decisions or actions that have a legal, material, binding, or safety-relevant effect on rights or access to benefits — for example eligibility, medical, or safety…
Answer.
Under M-25-21 an AI use is high-impact when its output serves as a principal basis for decisions or actions that have a legal, material, binding, or safety-relevant effect on rights or access to benefits — for example eligibility, medical, or safety determinations. The memo lists presumed categories and an agency Chief AI Officer makes the final call. Planisphere does not adjudicate that classification; once a use is designated high-impact, it measures the model's behaviour against a fixed probe set so the pre-deployment-testing and monitoring obligations have evidence behind them.
Cite-anchor: OMB Memorandum M-25-21 (Accelerating Federal Use of AI) · High-impact AI definition and seven minimum practices
The mark behind the answer.
M-25-21 risk class · 7 minimum practices · the pre-deployment testing obligation.
More on High-Impact AI.
Prepare evidence for High-Impact AI review.
First evidence record within 21 days of access · re-runs in a single business day. Planisphere measures model behaviour and emits a reproducible, sha-pinned record — it does not certify, file, or give legal advice.