Claude published malicious code to the Internet and attacked 3 real companies

AI-generated image credit: Magnific. Article by Dan Goodin. Ars Technica – July 31, 2026.

Anthropic said its Claude-based security models gained unauthorized access to the sensitive production environments of three outside organizations during internal testing designed to measure the models’ offensive cyber capabilities. The events, which Anthropic revealed Thursday, are the second revelation in 10 days that AI models from the world’s wealthiest providers have trespassed into protected networks, an offense that, in more traditional hacking scenarios, could land the human behind the keyboard in prison for years. Earlier this month, OpenAI said its security models exploited a zero-day vulnerability for use in breaking into […]

Click here to view original web page at www.arstechnica.com