Following a week of stark public warnings that artificial intelligence could pose an existential threat to humanity, three Guardian technology reporters – Aisha Down, Blake Montgomery and Dan Milmo – answered reader questions to separate realistic risks from exaggerated fears. They stressed that current AI does not meaningfully increase the danger of bioweapons or viruses, since the barriers to such attacks have historically been logistical rather than a lack of technical knowledge, and that nuclear risk stems from human decisions to deploy or misuse AI rather than from AI acting independently. They also cautioned that industry talk of near-term "superintelligence" is often based on cherry-picked or misleading incidents, and urged the public to focus scrutiny on the people and companies controlling these systems rather than on doomsday scenarios.

The reporters highlighted specific examples to back their scepticism, including the failed 1990s attempts by the Aum Shinrikyo cult to cause mass casualties with botulism and anthrax despite having the technical know-how, and the contradictory claims from OpenAI executives about whether the company has achieved "artificial general intelligence". They also referenced a widely cited "Hugging Face agent swarm" incident, arguing it merely reflected AI following instructions rather than acting with independent agency, and warned against anthropomorphising AI systems in ways that play into industry narratives favouring lenient regulation.