Anthropic Blocks Several AI Pathogen-Weaponization Attempts, Urges New DNA Controls
Updated
Updated · The New York Times · Sep 14
Anthropic Blocks Several AI Pathogen-Weaponization Attempts, Urges New DNA Controls
1 articles · Updated · The New York Times · Sep 14
Summary
Anthropic said Thursday it stopped several attempts to use its AI models in ways that may have been aimed at making pathogens more dangerous.
The disclosure adds concrete evidence to warnings from biosecurity experts that chatbots can provide guidance on enhancing pathogens and that AI is accelerating toxin and virus design.
Dario Amodei, Anthropic’s chief executive, followed with a Saturday call to slow AI development and explicitly ban using AI to produce biological weapons or letting users do so.
The warning lands after scientists in August reported using AI to build viruses not seen in nature, widening concern that the technology could lower barriers to biological warfare.