Anthropic says it blocked attempts by some scientists to use its artificial intelligence systems for work connected to biological weapons, according to Corriere della Sera. The Italian newspaper reported the development on Thursday, September 10, 2026.
The same reporting was picked up by several other outlets, including Yahoo Finanza, which said Anthropic stated its AI had aided possible biological weapons research. The coverage does not specify how many people were involved, which systems were used, or when the attempts took place.
The reports sit alongside a broader thread of recent coverage about AI and virus design. Tgcom24 reported that artificial intelligence has begun producing replicable viruses in the laboratory, and rivistastudio.com said an AI designed viruses that do not exist in nature and can be replicated in a lab. Mondosanità framed the same subject with the argument that the news may be better than it first appears.
Taken together, the reports point to a widening discussion about whether AI models can be misused in ways that touch on biosecurity, and about what role AI developers play in preventing that. Anthropic's statement, as relayed by Corriere della Sera, is one company's account of blocking such attempts rather than a public finding about any specific incident.
What to watch next: whether Anthropic or other AI developers release more detail on how these attempts were detected and stopped, and how biosecurity rules for AI systems develop in response.
