Anthropic CEO Dario Amodei has called for a "slowdown" of AI development to prevent it outrunning human control, amid fears that AI could be used to develop bioweapons. Brendan Walker-Munro, a security researcher at Southern Cross University, said AI is "lowering those knowledge barriers" that previously limited access to dangerous biological tools to qualified scientists.
David Heslop, a public health researcher at the University of New South Wales, said the speed of AI progress makes him "very concerned." "The tools are advancing at a rate which means that human oversight is becoming challenging," he said. An advanced AI model could provide the information required to make a bioweapon, potentially a virus, bacteria, fungus, or something never seen before. Heslop also warned that AI could help bad actors plan to deploy pre-existing viruses or toxins, though such work would still require expertise and expensive lab equipment. "We're not at the point, yet, where the LLM gives you a superpower," he said. "But it's certainly something that could be done in a laboratory."
David Kainer, a computational biologist at the University of Queensland, said AI tools applied to plant science could "potentially destroy large amounts of food production" and could be used to "engineer super-pathogens for crops."
On the positive side, researchers at Anthropic have announced that AI helped them discover a new class of enzymes that could represent a new gene editing mechanism. The article also notes that in 2022, researchers at Collaborations Pharmaceuticals demonstrated they could use their company's AI software to design dangerous biological agents, highlighting the dual-use nature of the technology.