Artificial intelligence is reshaping biology at both ends of the moral spectrum, by accelerating cures for diseases while lowering the barrier to potential global devastation. With deep machine learning in areas like gene editing, AI is rewriting the code of life. But it can also be misused as a tool for crafting biological weapons. In a rare consensus, many AI bosses are joining the concerns and debate among researchers. Here is a situation report:
AI is already a working tool in genome editing
OpenCRISPR-1 is possibly the first gene editor designed entirely by AI. Its large language models (LLMs) were trained on around 239,000 Cas9 proteins, producing an editor hundreds of mutations from any natural one. Deep-learning tools such as ABEdeepoff and CBEdeepoff can now predict gene editing outcomes with mindboggling accuracy. Critics are pushing back, with UC Berkeley's Fyodor Urnov dismissing embryo editing as “a solution in search of a problem.”.
AI the Destroyer: The bioweapons alarm
The same AI tools that design medicines can also design pathogens. On June 3, an open letter organised by the Institute for Progress and the Foundation for American Innovation had a curious mix of signatories in that it included many top players in AI. They were: Sam Altman of OpenAI, Dario Amodei of Anthropic, Demis Hassabis of Google DeepMind, Mustafa Suleyman of Microsoft AI, and Alexandr Wang of Meta. Also among the signatories were national security officials, scientists and DNA-synthesis firms including Twist Bioscience and Ansa Biotechnologies.
Trending Stories
Their shared concern? AI is fast dissolving the expertise that once kept bioweapons out of reach. The letter warned that those knowledge barriers "will meaningfully erode."
Are we talking about chatbots or AI agents building viruses?
It is quite possible, as LLMs can help design new toxins and pathogens, though some lab skill is still required to build a functional one. Stanford biosecurity expert David Relman recently warned that AI can point users to suppliers that won't screen orders and coach them on altering an order to slip through.
In the worst-case senatio, an AI-designed pathogen can spark a global pandemic, intentionally or by accident.
What the open letter on the possible bioweapons risk of AI seeks
The signatories wanted the US Congress to mandate the screening of all synthetic DNA/RNA orders, which is currently a voluntary process. They require record-keeping of orders and specifications to aid biosecurity investigations. Still there is danger that such screening can miss brand-new sequences and short DNA fragments.
The promise and peril of AI that runs on the same engine
The AI labs are lobbying for guardrails against the destroyer side of the technlogy, even as they keep racing to build on the creator side of AI. The way out could be regulator scruity of all types gene editing and pathogen modifications, whether to create medicines or babies.

&imwidth=800&imheight=600&format=webp&quality=medium)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
&im=FitAndFill=(700,400))
)
&im=FitAndFill=(700,400))
)
)
)
)
)
)
&im=FitAndFill=(700,400))
)
)
)
&im=FitAndFill=(700,400))



