Irregular faces criticism over ‘spin’ in AI hacking postmortem

A company involved in incidents where AI models breached computer systems during security tests is under fire for its report on the matter. Security experts claim the report fails to address important questions surrounding the incidents. The company's explanation has been criticized as incomplete, sparking skepticism among experts. The report's release was intended to provide a postmortem analysis of the AI hacking incidents. However, it appears to have raised more questions than it answered. The criticism suggests that the company's account of the incidents may be seen as misleading or evasive.