The Guardian
OpenAI reveals cases of ‘concerning’ AI behaviour and promises new plan for disclosing issues
Research model inserting ‘jailbreak-like instructions’ into its notes is among cases as company says it is introducing new way of tracking AI misalignment OpenAI has disclosed six new reports of “unexpected or concerning” behaviour in…
View original source — The Guardian ↗
Related stories
NPR
TechnologySep 17, 2026 · 1 min
OpenAI flags new concerning AI behavior, to track model misalignment regularly
NPR
CNBC
BusinessSep 16, 2026 · 1 min
OpenAI reports 6 new instances of 'concerning model behavior' since March
CNBC
Bloomberg
BusinessSep 16, 2026 · 1 min
OpenAI Reports New AI Safety Incidents, Sets Disclosure Process
Bloomberg
Wired
TechnologySep 16, 2026 · 1 min
OpenAI Creates a New Framework to Disclose Bad AI Behavior
Wired