The makers of ChatGPT, OpenAI, have issued a stark warning about the potential misuse of advanced artificial intelligence to aid in the creation of biological weapons.
In a recent blog post, the company cautioned that as AI becomes more adept in fields like biological research, its capabilities could be exploited by malicious actors to generate harmful information — including instructions for replicating known biological threats.
While AI is widely celebrated for its role in accelerating drug discovery and vaccine development, OpenAI emphasized that these same tools could be misused by individuals with technical knowledge. Although physical access to laboratories and biological agents remains a critical barrier, the company warned that such limitations are “not absolute.”
OpenAI’s head of preparedness, Johannes Heidecke, told Axios that future AI models likely won’t be able to build bioweapons on their own, but could enable less experienced users to recreate existing bio-threats.
“We are more worried about replicating things that experts already are very familiar with,” Heidecke said.
To mitigate the risks, OpenAI is investing in robust safety systems designed to detect and block dangerous queries. Heidecke stressed the need for “near-perfect” accuracy in detecting misuse, noting that anything less poses unacceptable risk.
The company is also working closely with experts in biosecurity, bioterrorism, and international health to design appropriate safeguards.
The warning comes amid broader concerns about AI misuse. Past bioterrorism events, such as the 2001 anthrax attacks in the United States, have highlighted the catastrophic potential of biological threats. Last year, leading scientists warned that future AI systems could be weaponized unless governments implement urgent regulatory frameworks.
OpenAI’s statement underscores the growing awareness within the tech industry of AI’s dual-use nature — as both a revolutionary tool and a potential vector for serious harm.
