← Field Notes
SEP 16 · Firsthand · via OpenAI ObservabilityOff-Brief Alert

OpenAI gives itself six business days to publish what a model did

A clock on disclosure is a design decision as much as a policy. Six days means the report ships before the fix, so the reader sees the behavior raw. Copy the clock, not the prose.

Read the source ↗ attached to Observability →
emerging

watching → emerging → established

Seen across products but the shape is still settling. Adopt with an exit: expect the anatomy to shift.

29 documented sightings since Aug 2026 — every one traces to a dated source.

See the evidence →

Machine summary of the source

OpenAI published a framework for tracking, investigating and disclosing model misalignment on September 16, with six reports beside it. Any employee can flag a case. Each case goes on one of three tracks. Ready for Disclosure is published within six business days. Minor Investigation gets twelve. Larger Investigation covers cases that involve third parties or security and has no fixed clock. Disputes go to the Safety Advisory Group, then to leadership. OpenAI says it will publish before it has explained or fixed a behavior, and that some cases will turn out to be spurious. It says serious incidents should also go to the US government, and it is drafting how.

The summary above is generated; the note at the top is the editorial judgment. Primary source ↗