MIT experts warn AI and biotech could create lethal pathogens

MIT Technology Review hosted a live Roundtables event on Wednesday that asked whether artificial intelligence could become an extinction‑level threat and how the technology might be weaponised as a bioweapon. The discussion brought together AI researchers, biosecurity experts and policy makers to examine the worst‑case scenarios and what can be done now to prevent them.

The debate comes at a time when AI systems are moving from niche tools to widely deployed products. Large language models such as OpenAI’s GPT‑4 and Google’s Gemini can generate convincing text, images and code, and they are being integrated into everything from search engines to medical diagnostics. At the same time, the bio‑tech sector has seen a surge in gene‑editing tools and synthetic‑biology kits that lower the barrier to creating harmful pathogens. Experts warn that the combination of powerful, easy‑to‑use AI and accessible biotech could accelerate the design of lethal viruses or toxins, making it harder for authorities to detect and stop malicious actors.

During the roundtable, former U.S. intelligence officer Michael C. Horowitz warned that “the speed at which AI can iterate designs outpaces traditional oversight.” AI researcher Kate Crawford highlighted the lack of robust safety testing for generative models, while biosecurity specialist Dr. James Giordano pointed to recent incidents of lab‑breach reports as evidence that the system is already fragile. The panel cited a 2023 study that estimated the cost of creating a synthetic virus could drop from millions to a few thousand dollars with AI assistance. Several participants called for immediate international norms, tighter export controls on dual‑use AI tools and a coordinated research agenda on AI‑driven bio‑risk.

The next steps will likely involve governments and industry groups drafting new regulations. The U.S. National Security Commission on AI is expected to release a report later this year that includes recommendations on bioweapon safeguards. In Europe, the European Commission plans a summit on AI and biosecurity in early 2025. The MIT Technology Review roundtables will continue quarterly, providing a forum for scientists and policymakers to track progress and adjust strategies as the technology evolves.