Anthropic Blocked Biological Weapon Attempts After Reviewing Safety Reports

Anthropic has confirmed it blocked several attempts this year to build biological weapons with its artificial intelligence systems. The tech giant arrived at this conclusion after reviewing reports that scientists were using their platforms for research capable of creating dangerous pathogens. A document obtained by the New York Times revealed these findings without naming the specific researchers or institutions involved. Anthropic could not tell if the work served a legitimate purpose or something nefarious, so they chose caution over risk. Valid biological inquiry can lead to breakthroughs like vaccines, yet that same line of study carries the potential for harm.

The company's CEO Dario Amodei co-founded the firm and signed off on these safety protocols. The report published Thursday also highlighted other misuses of their AI models like chatbot Claude over the past eight months. China, Russia, and Yemen tried to use Claude to design software for firearms, missiles, armed drones, and bombs according to the investigation. Suspected Chinese and Iranian government actors targeted dissident and diaspora communities for surveillance using these tools. Russian state media generated election propaganda with the help of the chatbot, including fabricated claims about parliamentary elections in Moldova. These actions show how readily available technology can be weaponized by various global powers.