Listen again · In Conversation
Technology and AI: What Could Change in Bioweapons Development? October 15, 2026 · 3:51 · With Maya and Leo
I’m Maya, and Leo is our columnist for technology and digital life. Thanks for listening. Leo, the question today is whether artificial intelligence could change the risk of bioweapons.
Transcript Tap a sentence to hear it from there.
Maya Welcome to Slow English FM. This is Technology and AI, where we look at how new tools may affect the world around us. I’m Maya, and Leo is our columnist for technology and digital life. Thanks for listening. Maya Leo, the question today is whether artificial intelligence could change the risk of bioweapons. What warning has brought this issue into focus? Leo The Guardian reports that Sir Ken McCallum, head of the UK intelligence service MI5, warned in an annual speech that AI could “change the stakes” for existing and new bioweapons. That is a warning about what might happen. It is not proof that AI has already enabled one. Maya What kinds of help are people concerned AI might provide? Leo The Guardian describes two broad possibilities. A chatbot might help someone find information, plan research or work through problems. Separately, a well-resourced group that already has skilled scientists and specialist facilities might use AI to support its work. These are potential uses, not evidence that the technology has produced a bioweapon. Maya So the concern is not only that someone could get information. What still makes this kind of work difficult? Leo The Guardian quotes science and security researcher Filippa Lentzos, who says information is only one part of the problem. People would also need biological materials and equipment, practical know-how, and ways to test and carry out the work. The article describes those as substantial barriers. In other words, access to information does not by itself mean someone can put a dangerous plan into practice. Maya The article also discusses a suspected plot in India. What does it say, and what remains uncertain? Leo The Guardian reports that police uncovered a suspected bioterrorism plot involving a doctor and several others. The people allegedly drew on chatbot queries and AI-driven search, and the report says this was to help them make ricin, a deadly toxin. But the article says it is unclear whether the AI responses helped. So this case raises a concern, but it does not establish that AI made the plot possible. Maya And there is also a report from the AI company Anthropic. How should listeners understand that? Leo The Guardian says Anthropic reported five cases involving scientists in unnamed countries who tried to use its Claude chatbot for work that may have involved viruses or toxins. That is the company’s account. The article does not present those cases as independently established bioweapons work, and it keeps the nature of the work uncertain. Maya What is the clearest way to sum up the issue, then? Leo The warning is that AI could make some parts of dangerous research easier, for different kinds of users. But the evidence described by The Guardian does not show that AI has removed the other major barriers or that it has, on its own, enabled a bioweapon. My interpretation is that both points matter: the possible risk deserves attention, but it should not be confused with a proven outcome. Maya Leo, thank you. And thanks to you for listening to Slow English FM.
Slow English FMThe world in clear English.
--:-- Local time
