Is regulation enough to stop AI systems killing us all? | Letters

Martyn Thomas wonders how any rational person can argue that independent oversight and regulation will be sufficient Once again a frontier AI model has been withdrawn because it fails safety tests ( OpenAI scraps release of new model over safety concerns in internal testing, 28 September ). There is the usual call for “independent oversight and regulation”, though never any explanation of how this would work or, most importantly, what evidence should persuade a regulator that an AI system is adequately safe. There is a long engineering tradition of regulating safety-critical software systems, such as those controlling aircraft or nuclear power stations; international standards require “safety cases” – rigorous analyses that show that the possible hazards have been identified and controlled and that the probability of an accident is extremely low. Continue reading...

AI systems pose existential risk

This is an opinion letter arguing regulation is inadequate, using one model withdrawal as evidence. It raises a legitimate structural concern about missing rigorous safety standards but offers no new data, only argument. Score reflects concern about governance gaps rather than a concrete harm.

The author expresses strong skepticism about the effectiveness of regulation to prevent catastrophic AI failures, citing a recent model withdrawal due to safety concerns. The core implication is that current regulatory approaches are insufficient to prevent AI systems from posing an existential threat.

The text discusses the potential danger of AI systems and the inadequacy of regulation. It mentions a specific incident of an AI model failing safety tests. The overall tone is dystopian, but the evidence is speculative.

SOURCE     The Guardian: Artificial Intelligence (tier 1)
   ↓
DOCUMENT   62a4fd2b-534c-4bb4-b2cc-b341bedce99b
           https://theguardian.com/technology/2026/oct/06/is-regulation-enough-to-stop-ai-systems-killing-us-all
   ↓
EVIDENCE   2 extracted excerpts
   ↓
MODEL RUN  4 runs, methodology 1.0.1
   ↓
SCORE      -67  (Strongly adverse)
   ↓
CONFIDENCE 51%