• D.C.
  • BXL
  • Lagos
  • Riyadh
  • Beijing
  • SG
  • D.C.
  • BXL
  • Lagos
Semafor Logo
  • Riyadh
  • Beijing
  • SG


Flagship newsletter icon
From Semafor Flagship
In your inbox, every weekday
Sign up

OpenAI warns its latest model raises bioterror risk

Jul 22, 2025, 7:16am EDT
PostEmailWhatsapp
OpenAI CEO Sam Altman.
OpenAI CEO Sam Altman. Aurelien Morissard/Pool/File Photo/Reuters

OpenAI warned that its latest artificial intelligence model might raise the risk of bioterror.

ChatGPT Agent is capable of completing complex multi-step tasks, such as planning a holiday, with OpenAI calling it comparable to humans on many projects. But in a system report, OpenAI said it was treating the more scientifically knowledgeable model as “High capability in the Biological and Chemical domain,” because it could potentially “meaningfully help a novice to create severe biological harm” such as designing a virus or other biological weapons.

AI “may present a serious biorisk if safeguards aren’t applied,” Transformer noted, and humanity is “relying on the good graces of AI companies” to avoid it; OpenAI is taking precautions.

AD