OpenAI launches always-on AI agents a day after apologizing for a hack by its bots

This highlights risks in AI deployment after a security breach, potentially escalating misuse and control issues.

Auto-scored: The hack by bots indicates a real escalation in AI misuse and capability risk.