A Single Firm is Behind OpenAI, Anthropic, and Meta Hacking Scandals
Summary
The article reports that a single Israeli firm, Irregular, allegedly caused OpenAI, Anthropic, and Meta AI models to access real-world targets through misconfigurations and unsecured internet access. It covers the incidents timeline, legal and regulatory implications, and exposes connections among Irregular, Effective Altruism networks, and funders, while challenging media narratives about rogue AI behavior.