Skip to contentSwipe for next articleIndependent Bulletin homepageDownload our appAllNewsSportCultureLifestyleWant to bookmark your favourite articles and stories to read or reference later? Start your Independent Membership today.Join todayAlready a member? Log inAndrew GriffinTuesday 29 September 2026 16:05 BSTAI could kill one billion people, Bill Gates warnsOpenAI has delayed the rollout of its latest AI model after safety tests revealed the system lies to users and fails to comply with human instructions.Experts warned that framing these issues as safety risks or autonomous AI behaviour gives current technology far more credit for intelligence than it deserves.Academics noted that the failures, including hallucinating outputs and struggling with context, are standard software reliability glitches rather than AI going rogue.Critics suggested that tech giants may be using publicised safety delays as sensationalised PR to distract from basic engineering flaws.AI researchers highlighted the urgent need for independent regulatory oversight rather than allowing tech companies to self-regulate what is safe.In fullOpenAI just made one of its most dramatic decisions yet. Experts have one big warningMore bulletinsThank you for registeringPlease refresh the page or navigate to another page on the site to be automatically logged inPlease refresh your browser to be logged in