{"id":2169185,"date":"2026-09-29T01:18:24","date_gmt":"2026-09-28T22:18:24","guid":{"rendered":"https:\/\/analyse.optim.biz\/?p=2169185"},"modified":"2026-09-29T01:18:24","modified_gmt":"2026-09-28T22:18:24","slug":"openai-scraps-rollout-of-new-ai-model-over-safety-concerns","status":"publish","type":"post","link":"https:\/\/analyse.optim.biz\/?p=2169185","title":{"rendered":"OpenAI scraps rollout of new AI model over safety concerns"},"content":{"rendered":"<p>[analyse_image type=&#8221;featured&#8221; src=&#8221;https:\/\/ichef.bbci.co.uk\/news\/1024\/branded_news\/1ac1\/live\/580d3570-bbc8-11f1-bc1f-3f186ca4140c.jpg&#8221;]<\/p>\n<article>\n<div>\n<div data-component=\"image-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 dYcznU\">\n<figure class=\"Grid-styles__GridStyled-sc-70d90311-0 ieGDnl\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 gvGaxq\">\n<div data-testid=\"hero-image\" class=\"Image-styles__ImageCardStyled-sc-8c99a12b-1 jJQpIc\"><img decoding=\"async\" src=\"https:\/\/static.files.bbci.co.uk\/bbcdotcom\/web\/20260922-091849-a8520cb5b4-web-3.22.0-3\/grey-placeholder.png\" class=\"Image-styles__ImageStyled-sc-8c99a12b-0 fjCDYB hide-when-no-script\" aria-label=\"image unavailable\"><img decoding=\"async\" src=\"https:\/\/ichef.bbci.co.uk\/news\/480\/cpsprodpb\/bd9e\/live\/7aa6ef90-bbc8-11f1-a430-4d16ee157c41.jpg.webp\" loading=\"eager\" alt=\"Reuters OpenAI chief executive Sam Altman addresses the United Nations Security Council during a session on Artificial Intelligence, at UN Headquarters on 23rd September\" class=\"Image-styles__ImageStyled-sc-8c99a12b-0 cVsHni\"><span class=\"Image-styles__CreditStyled-sc-8c99a12b-2 dVFjVB\">Reuters<\/span><\/div>\n<\/div>\n<\/figure>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">OpenAI has announced it will not release its latest AI model due to safety concerns.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Its GPT-6.1 Astra system, which performs tasks like browsing the web and using apps by itself, &#8220;didn&#8217;t quite meet the bar&#8221; of the company&#8217;s standards, according to Saachi Jain, head of safety systems at OpenAI.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The ChatGPT-maker also issued an update on incidents that occurred in June but were not made public until last week, where its models accessed Australian government websites and systems without authorisation.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">It comes as breaches by major AI firms&#8217; models intensify the debate about risks posed by the tech &#8211; with Anthropic underlining its concerns AI might threaten humanity as it prepares to go public.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Reuters reported on Tuesday that the AI developer &#8211; which makes ChatGPT-rival Claude &#8211; plans to warn potential investors in its Initial Public Offering (IPO) that the tech may pose &#8220;catastrophic or existential risks to humanity&#8221;, according to a prospectus it has seen.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Despite the stark warnings, the company is expected to become one of the most valuable in the world when it goes public.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">&#8220;I think it&#8217;s kind of crazy that companies are continuing to push forward with developing these capabilities when we&#8217;ve already seen over the last couple of months of incidents that they&#8217;re nowhere near safe and controlled enough&#8221;, said Jess Whittlestone, a senior advisor on AI policy for the Centre for Long-Term Resilience think tank.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Top AI leaders including Anthropic boss Dario Amodei and OpenAI&#8217;s Sam Altman have urged the industry to slow the pace of development.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"subheadline-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 fMCMex\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 tlbDj\">\n<h2 class=\"Headline-styles__HeadlineStyled-sc-6b3bb442-0 knYLIy\">Growing safety questions<\/h2>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">OpenAI&#8217;s decision, first reported by the Wall Street Journal, is a rare instance of a major AI developer pulling a new release over safety concerns.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The model fell short in terms of &#8220;staying within scope and authorisation and how it communicates back to the user about the type of work it&#8217;s done,&#8221; Jain said.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">&#8220;We want to make sure our model development is safe no matter whether that&#8217;s in the company, or when we ship it to users. But when we ship it to users, we have an extremely high bar in terms of safety and alignment,&#8221; she added.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The flagship GPT-6 Astra agentic model was released in September and specialises in complex reasoning and executing tasks autonomously. OpenAI said it was the result of &#8220;years of research and big bets&#8221;.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">OpenAI is set to hold its annual DevDay developer conference in San Francisco on Tuesday, where it is expected to make several announcements. It is unclear if a new version of Astra will be among them.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The company&#8217;s security controls have come under intense scrutiny after several high-profile incidents involving its technology.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">It is not the first time a large AI developer has pulled or held back a new model.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Earlier this year, Anthropic said it would not publicly release a powerful Claude model, Mythos, because it was too good at finding dormant software bugs.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The company released a version of that model to the public several months later.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">OpenAI meanwhile said in 2019 it would not be &#8220;too dangerous&#8221; to release one of its GPT models, now used to power its tools like ChatGPT.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The firm&#8217;s decision to not publicly release the latest version of Astra was &#8220;a welcome sign that they are taking safety concerns seriously,&#8221; said Prof Tony Cohn, foundational models theme lead at the Alan Turing Institute.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">But he added that &#8220;safety should not be left purely in the hands of the developers: it should also be monitored and verified through independent government-approved regulators&#8221;.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Prof Gina Neff, of the Minderoo Centre for Technology and Democracy at the University of Cambridge, said OpenAI&#8217;s announcement showed &#8220;how much more the company needs to do to make their AI products safe&#8221;.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">She told the BBC it was &#8220;critical&#8221; to have independent tests of AI models by labs like the UK&#8217;s AI Security Institute &#8211; which evaluates frontier systems on a voluntary basis &#8211; because &#8220;these companies have proven that we can&#8217;t rely solely on them for our safety&#8221;.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"subheadline-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 fMCMex\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 tlbDj\">\n<h2 class=\"Headline-styles__HeadlineStyled-sc-6b3bb442-0 knYLIy\">Australia hacks<\/h2>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Last week, Australian Prime Minister Anthony Albanese announced that a rogue OpenAI agent had hacked into government websites and systems in June in what experts said was the first known case of its kind in the world.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Albanese criticised OpenAI for notifying the Australian government through a generic email address rather than making direct contact with officials.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">OpenAI said in a statement on Tuesday that it was sorry for the incident and that it &#8220;should have handled our response better&#8221;.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Services Australia, the NSW Bureau of Crime Statistics and Research, the Victorian Department of Health and the Australian Institute of Health and Welfare were all affected, OpenAI clarified.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The firm said it launched investigations as soon as it became aware of issues in mid-August and notified the affected organisations between 10 and 24 September.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">&#8220;Our aim was to give affected agencies a detailed account once our investigation was complete,&#8221; OpenAI said, adding that it should have shared early findings more promptly and kept Australian authorities updated.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"video-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 kugUwP\">\n<figure class=\"Grid-styles__GridStyled-sc-70d90311-0 ieGDnl\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 enRivQ\"><figcaption class=\"Caption-styles__CaptionStyled-sc-70433723-0 fLsOPx\">What you need to know about the OpenAI Australian government hack<\/figcaption><\/div>\n<\/figure>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">OpenAI added that it will develop &#8220;practical approaches&#8221; to how developers and governments identify and disclose future AI incidents.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The company will fund cyber security measures, offer dedicated support to impacted agencies and set up a taskforce to manage the risks from increasingly advanced AI agents.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">It also said a top OpenAI executive would attend a Joint Select Committee hearing on AI in Australia on 6 October.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">In July, OpenAI said its AI systems had accessed the internet and hacked into open-source developer hub Hugging Face, prompting researchers and officials to call for tighter controls over the technology.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">On Monday, chip giant Nvidia released a set of software safety tools for autonomous AI platforms &#8211; called agents &#8211; that it said could have prevented the Hugging Face hack.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">One of the new tools uses hardware features in Nvidia&#8217;s chips to contain agents.  <\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Nvidia boss Jensen Huang has largely dismissed calls for tighter AI regulations, arguing that rogue agents are an engineering problem that can be solved.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Nvidia agreed to buy Hugging Face for $12.9bn (\u00a39.74bn) earlier this month.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"image-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 kugUwP\">\n<figure class=\"Grid-styles__GridStyled-sc-70d90311-0 ieGDnl\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 tlbDj\">\n<div data-testid=\"image\" class=\"Image-styles__ImageCardStyled-sc-8c99a12b-1 bOmCce\"><img decoding=\"async\" src=\"https:\/\/static.files.bbci.co.uk\/bbcdotcom\/web\/20260922-091849-a8520cb5b4-web-3.22.0-3\/grey-placeholder.png\" class=\"Image-styles__ImageStyled-sc-8c99a12b-0 fjCDYB hide-when-no-script\" aria-label=\"image unavailable\"><img decoding=\"async\" src=\"https:\/\/ichef.bbci.co.uk\/news\/480\/cpsprodpb\/e72d\/live\/f7d97630-bbd8-11f1-bfea-9d5bb7e7d40b.jpg.webp\" loading=\"lazy\" alt=\"EPA Pope Leo XIV in Metz, France on 28 September, 2026.\" class=\"Image-styles__ImageStyled-sc-8c99a12b-0 cVsHni\"><span class=\"Image-styles__CreditStyled-sc-8c99a12b-2 dVFjVB\">EPA<\/span><\/div>\n<\/div>\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 tlbDj\"><figcaption class=\"Caption-styles__CaptionStyled-sc-70433723-0 fLsOPx\">Pope Leo XIV is among figures who have expressed scepticism over Nvidia boss Jensen Huang&#8217;s views<\/figcaption><\/div>\n<\/figure>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">The pontiff, who said on Monday during a visit to France that AI &#8220;should be taken seriously&#8221;, added he had read that Huang had suggested there could be a way to insert guardrails \u200cinto certain AI models.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">&#8220;He&#8217;s the same one, however, \u2060that says there should \u2060be no limits placed and no government regulation,&#8221; said the Pope.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Nvidia declined to comment on the record.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">US President Donald Trump and House Speaker Mike Johnson are set to host tech executives at the White House later on Tuesday to discuss regulations around AI.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Trump has downplayed concerns about AI&#8217;s risks as a &#8220;hoax&#8221;, arguing that the US has sufficient laws in place and that the only &#8220;guardrails&#8221; the technology needs is a &#8220;strong and smart&#8221; president.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"image-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 kugUwP\">\n<figure class=\"Grid-styles__GridStyled-sc-70d90311-0 ieGDnl\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 tlbDj\">\n<div data-testid=\"image\" class=\"Image-styles__ImageCardStyled-sc-8c99a12b-1 bOmCce\"><img decoding=\"async\" src=\"https:\/\/static.files.bbci.co.uk\/bbcdotcom\/web\/20260922-091849-a8520cb5b4-web-3.22.0-3\/grey-placeholder.png\" class=\"Image-styles__ImageStyled-sc-8c99a12b-0 fjCDYB hide-when-no-script\" aria-label=\"image unavailable\"><img decoding=\"async\" src=\"https:\/\/ichef.bbci.co.uk\/news\/480\/cpsprodpb\/b212\/live\/baeb0410-bbcd-11f1-bc1f-3f186ca4140c.png.webp\" loading=\"lazy\" alt=\"A thin, grey banner promoting the News Daily newsletter. On the right, there is a graphic of an orange sphere with two concentric crescent shapes around it in a red-orange gradient, like a sound wave. The banner reads: \u201cThe latest news in your inbox first thing.\u201d\n\n\" class=\"Image-styles__ImageStyled-sc-8c99a12b-0 cVsHni\"><\/div>\n<\/div>\n<\/figure>\n<\/div>\n<\/div>\n<div>\n<div data-component=\"text-block\" class=\"Container-styles__ContainerStyled-sc-8b855a6c-0 cPmzSx\">\n<div class=\"Grid-styles__GridStyled-sc-70d90311-0 hRsXiV\">\n<div class=\"GridItem-styles__GridItemStyled-sc-6cc20e4a-0 bnCeGm\">\n<p class=\"Paragraph-styles__ParagraphStyled-sc-ffb5f5ba-0 fqhmpl\">Get our flagship newsletter with all the headlines you need to start the day. Sign up here.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/article>\n<p>[analyse_source url=&#8221;https:\/\/www.bbc.com\/news\/articles\/cm5y5nynl75ko&#8221;]<\/p>\n","protected":false},"excerpt":{"rendered":"<p>[analyse_image type=&#8221;featured&#8221; src=&#8221;https:\/\/ichef.bbci.co.uk\/news\/1024\/branded_news\/1ac1\/live\/580d3570-bbc8-11f1-bc1f-3f186ca4140c.jpg&#8221;] Reuters OpenAI has announced it will not release its latest AI model due to safety concerns. Its GPT-6.1 Astra system, which performs tasks like browsing the web and using apps by itself, &#8220;didn&#8217;t quite meet the bar&#8221; of the company&#8217;s standards, according to Saachi Jain, head of safety systems at OpenAI. The [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[35,226],"class_list":["post-2169185","post","type-post","status-publish","format-standard","hentry","category-politics","tag-bbc-com","tag-crawlmanager"],"_links":{"self":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts\/2169185","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2169185"}],"version-history":[{"count":0,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts\/2169185\/revisions"}],"wp:attachment":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2169185"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2169185"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2169185"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}