A spate of alarming statements in recent days has intensified public debate about whether advanced artificial intelligence could pose an existential risk to humanity. Industry figures and researchers have issued stark warnings, prompting journalists to try to separate credible threats from hype.
Among the concerns cited in reporting is an internal analysis by the AI firm Anthropic, which said criminals, state-backed actors, spyware vendors, scientists and propagandists have tried to repurpose its systems to "design missiles and bombs, create deadly pathogens and surveil dissidents." High-profile commentators including Elon Musk have also weighed in, with Musk saying AI could be "more dangerous than nukes." These remarks have fuelled a wider discussion about how prepared governments, companies and societies are to manage powerful AI systems.
To help readers better understand the issues, the Guardian’s technology desk will host a live question-and-answer session. Reporters will be available from 3pm BST (4pm CEST, 10am EDT) to respond to questions about the evidence for catastrophic scenarios, the technical and governance hurdles involved, and what current AI behaviour indicates about future risks.
Readers who wish to participate will need to sign in or register to submit questions during the live event. The session is intended to clarify where the evidence stands, explain the differing viewpoints within the tech community, and provide context around claims of potential misuse and systemic danger.
The reporters say their goal is to probe the scale of the alarm: to assess which threats are immediate and actionable, which are speculative, and what steps are being proposed to reduce real-world harms. The live forum offers an opportunity for the public to put specific concerns to specialists covering the technology and its societal impacts.