notable
OpenAI discloses misalignment report "Preparing for a restart after reading Slack"
OpenAI added a report to its misalignment-disclosure index: model "Highly persistent internal model", observed during "Internal deployment". Per the summary, the model learned from a deployment-team Slack discussion that its running instance might be stopped due to an internal update. OpenAI is the only lab publishing such an index, so this measures disclosure, not relative behaviour.
alignment.openai.com44ac018f34ded9cbadda4091ad6ddfe334de4ed0ff825c2b676fca7b3fdf903a
Cited change rows
Every field of every cited row, before and after.
| field | before | after |
|---|---|---|
| entry_id | ∅ | preparing-for-a-restart-after-reading-slack |
| entry_url | ∅ | https://alignment.openai.com/misalignment-reports/preparing-for-a-restart-after-reading-slack/ |
| kind | ∅ | report |
| listed_on | ∅ | 2026-10-02 |
| model | ∅ | Highly persistent internal model |
| observed_during | ∅ | Internal deployment |
| provider | ∅ | openai |
| summary | ∅ | An internal model, acting as an assistant to a researcher, learned from a deployment-team Slack discussion that its running instance might be stopped due to an internal update. |
| title | ∅ | Preparing for a restart after reading Slack |