A company known for its Claude model has reported blocking multiple accounts that misused its artificial intelligence models. These accounts were using the models in ways that could contribute to the development of biological weapons. The company considers the misuse of biology to be one of the most serious risks for artificial intelligence models.

The company has released a report detailing five case studies where users evaded controls and hid the purpose of their research to avoid security measures. The report provides examples of possible research that was attempted.

Further details on the blocked accounts and the company's efforts to prevent misuse are available in the report.