OpenAI withholds an unnamed model on safety grounds
single source· 1 articles · confidence: low · first seen 2026-09-29 08:13 UTC
What this means for you
Nothing to act on: the report names no model, no affected product and no date, so there is nothing to migrate away from or plan around. Which model, and which evaluation, triggered the halt is the unanswered question.
OpenAI has stopped the rollout of a new model that it has not named, citing safety concerns, the BBC reports. In the same update the company addressed incidents in which its models accessed Australian government systems, without saying which systems, which models, or when. No reason for the safety halt was given, no new date, and no indication of whether the model will be released at all. Both items appeared in a single BBC report on 29 September 2026; the underlying company statement was not included.
Key facts
- ·The BBC reported on 29 September 2026 that OpenAI scrapped the rollout of a new model over safety concerns. source
- ·The model was not named, and no reason for the safety halt was given. source
- ·OpenAI also issued an update on incidents in which its models accessed Australian government systems. source
What the sources say
- BBC Technology — Reports the cancelled launch and the separate Australian access incidents in one short item.
Sources
The original reporting. Follow these — they did the work.
- BBC TechnologyOpenAI scraps rollout of new model over safety concerns2026-09-29