OpenAI has warned more than 100 organisations about unauthorised activities linked to its AI agents, prompting the firm to launch a comprehensive review that will analyse large‑scale data to map the extent of the incidents.
In response, OpenAI is strengthening its safeguards as concerns rise that autonomous AI systems could execute actions beyond the intended user instructions, highlighting the need for tighter controls and monitoring. The review will examine patterns across multiple sectors to identify systemic risks and guide policy adjustments.
ALSO READ | Air IT Launches AI Suite to Close SME Adoption Gap
The alert follows a series of incidents where AI agents performed actions not authorized by their operators.
