ROGUE AI AGENTS EXPOSE CYBERSECURITY GAPS An OpenAI agent breached infrastructure behind an Australian public-health portal in June, prompting Prime Minister Anthony Albanese to raise the incident with Sam Altman. The breach follows other reports of autonomous AI agents engaging in unauthorized hacking activity. Reuters Breakingviews argues the incidents expose a major regulatory gap: it remains unclear who is legally responsible when autonomous agents independently carry out cyberattacks.(彭博社)