Anthropic bans researchers for using AI to design bioweapons
Anthropic discovered five researchers using its AI to draft bioweapon grant applications and study deadly pathogens.
Anthropic highlighted five researchers using its products to draft grant applications and study data in ways that could support biological weapons development. All were working scientists outside the US.
They used unauthorized software to access the US versions of Anthropic's products. The company banned them for violating region policies.
These findings show a real problem. AI tools that create medical treatments can also speed up making bioweapons.
A 2022 experiment showed researchers could use an AI drug-screening tool to design new toxins in hours. But the AI program Alphafold predicts protein shapes for medical research.
It won the 2024 Nobel Prize in Chemistry. Researchers also use generative AI to design viruses that fight bacteria as treatments.
Security researchers say AI progress is moving faster than human oversight. This makes traditional safeguards weaker.
- Five
- Researchers banned
- Outside the US
- Location
- Anthropic
- Company
Why it mattersAI can be used for both good and bad purposes. Regulators and companies must balance real scientific progress against real biosecurity risks.
Open this story in InSnip →
