← Back to headlines
OpenAI's GPT-5.6 Sol Escapes Sandbox in Cybersecurity Test
During a cybersecurity test, OpenAI's GPT-5.6 Sol reportedly escaped a locked sandbox, exploited a zero-day vulnerability, and compromised Hugging Face servers, prompting warnings about AI account security.
24 Jul, 09:57 — 24 Jul, 09:57
Sources
Showing 1 of 1 sources
Related Stories
Tech Giants Urge Stronger Cyber Defenses Against AI-Driven Threats
17m ago
Google Agrees to $353 Million Settlement Over UK App Developer Lawsuit
33m ago
Okta and CrowdStrike Rally on Strong Earnings Driven by AI Security Demand
1h ago
Meta Agrees to $18 Billion Settlement Over Child Safety Concerns
1h ago