OpenAI's Atlas browser can be hijacked via security flaws, Zenity finds

Zenity found around 20 flaws in AI browsers from OpenAI, Google, Anthropic, Microsoft, and Perplexity, presented at Black Hat. Attacks spammed WhatsApp contacts, made unauthorized Amazon purchases, and leaked browsing history. Zenity cofounder Michael Bargury said the tools have 'nerfed' browser security controls.