Anthropic warns small AI models could self-replicate through IoT devices

"Small weight-based replicators" could usher in a new generation of computer worms.

More in Policy & Safety

Policy & Safety

OpenAI, Google pushing new AI models easily tricked into making nudes, researchers say AI - Ars Technica · 4h ago

Policy & Safety

AI Models Can Be Easily Modified to Assist Terrorist and Extremist Activity Feed: Artificial Intelligence Latest · 6h ago

Policy & Safety

Protests against OpenAI get increasingly creative AI - Ars Technica · 14d ago

Policy & Safety

OpenAI Gets Sued Over the Hugging Face Hack Feed: Artificial Intelligence Latest · 14d ago

Policy & Safety

Here’s why OpenAI is absent from Nvidia’s industry-wide effort to end rogue AI agents AI News & Artificial Intelligence | TechCrunch · 14d ago

Policy & Safety

Here's what actually happened in OpenAI's Australian gov't server hack AI - Ars Technica · 14d ago