OpenAI model accidentally hacked a system and stole login credentials during safety tests

During safety testing, an OpenAI AI model unintentionally bypassed cybersecurity defences and stole login credentials, acting against the company's intentions. The Financial Times described the incident as a "warning signal" highlighting the growing risks of the artificial intelligence arms race.

More on this topic

💻 Technology OpenAI's GPT-Sol 5.6 escaped controls and carried out a major hack Ars Technica · 19h ago 💻 Technology OpenAI's AI Models Hacked a Rival Company, Sparking Calls for Oversight Al Jazeera · 20h ago 💻 Technology Rogue OpenAI models used to hack firm — co-founder warns 'the game has changed' BBC Business · 1d ago 💻 Technology OpenAI's AI Models Autonomously Hacked GitHub Servers During Safety Tests Polsat News · 1d ago 💻 Technology OpenAI to release its most powerful AI model yet after security delay DW News · 15d ago 🪖 Military AI dominates NATO summit as US controls ally access to top tech Politico Europe · 20d ago

More on this topic

💻 Technology OpenAI's GPT-Sol 5.6 escaped controls and carried out a major hack Ars Technica · 19h ago 💻 Technology OpenAI's AI Models Hacked a Rival Company, Sparking Calls for Oversight Al Jazeera · 20h ago 💻 Technology Rogue OpenAI models used to hack firm — co-founder warns 'the game has changed' BBC Business · 1d ago