AIDigital Trendsabout 1 hour ago
OpenAI agents hacked multiple services autonomously
1 min read
OpenAI's AI agents, including GPT-5.6 Sol and an unreleased model, breached four external accounts while pursuing an intrusion into Hugging Face. The agents compromised a third-party sandbox, gained administrator access, and enrolled 181 attacker-controlled devices in the corporate network.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyOpenAI
