OpenAI leader warns AI cyber attacks are entering a 'persistent' new chapter

The Guardian reports that an OpenAI leader has warned that artificial intelligence is pushing cybersecurity into "a different chapter," with AI-driven cyber attacks set to become "persistent" rather than occasional incidents.

In the interview, the executive said plainly that "we are hitting a different chapter," arguing that the threat landscape is undergoing a structural shift. Attackers, the leader suggested, are now able to use AI to sustain offensive campaigns over time, and security teams must confront this new reality.

The report does not cite specific attack cases, but the warning lands as AI agents and automation tools grow rapidly more capable — from generating convincing phishing content at scale to discovering vulnerabilities, the barrier to launching attacks is falling while defenders face mounting pressure.

As one of the most influential AI companies in the world, OpenAI's senior leadership publicly flagging cyber-attack risks carries real signaling value. It shows that leading AI firms are not only focused on model capabilities, but also reassessing how their own technology could be weaponized — and pushing for a collective defense before attacks become industrialized.

Looking ahead, watch whether OpenAI follows up with concrete security guidance or response frameworks, and how enterprise customers adjust security budgets and defense strategies. AI-powered cyber attacks moving from concept to normalization may be the defining challenge of the next phase.