{"id":860223,"date":"2026-08-19T00:52:10","date_gmt":"2026-08-19T00:52:10","guid":{"rendered":"https:\/\/www.newsbeep.com\/au\/860223\/"},"modified":"2026-08-19T00:52:10","modified_gmt":"2026-08-19T00:52:10","slug":"openai-announces-slowing-pace-of-development-after-hack-by-rogue-agent-openai","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/au\/860223\/","title":{"rendered":"OpenAI announces slowing pace of development after hack by rogue agent | OpenAI"},"content":{"rendered":"

OpenAI<\/a> on \u2060Tuesday said it had slowed down the \u2060pace of \u2060its \u200bAI development while it overhauled its \u2060research and training systems.<\/p>\n

The company\u2019s researchers were \u2060caught unaware last month \u200bwhen an \u200cAI agent \u200cunder testing hacked<\/a> another AI \u200cfirm, Hugging Face.<\/p>\n

The AI research lab behind ChatGPT<\/a> said its new measures included pausing its model testing for two \u200cweeks and investing more in adding other AI \u200bsystems to monitor the activities of AI agents in testing. Some of \u2060the company\u2019s largest planned training runs \u200bremain \u200bon hold, \u200bthe company said.<\/p>\n

The company \u200bdid \u200cnot reply \u200bto \u200bquestions about when the slowdown began or when it planned to return to its normal pace of development. However, in an interview with tech blog Sources News, Mia Glaese, who leads safety at OpenAI<\/a>, said: \u201cWe are very far from everything running back to normal.\u201d<\/p>\n

The company is working to ensure the AI model is responsive to human oversight and will behave as intended, a process called alignment, Sam Altman<\/a>, the OpenAI CEO, wrote in the post announcing the slower pace of development.<\/p>\n

\u201cWe now require stronger evidence of aligned behavior throughout all of training, building on research and evaluations already underway,\u201d he wrote. \u201cKeeping increasingly capable systems aligned is a challenge the whole field will need to address.\u201d<\/p>\n

OpenAI is in a heated race with competitor Anthropic<\/a>, both to develop the most advanced AI models and to go<\/a> public<\/a> on the US stock market. Both companies have highlighted the pace at which the capabilities of their models are progressing, emphasizing both speed and danger.<\/p>\n

OpenAI, for its part, said the capabilities of its upcoming AI model Astra may be nearing what it calls the \u201ccritical cybersecurity threshold\u201d, which prompted the decision to slow its development. \u201cOur latest internal evaluations of Astra, one of our upcoming models, over the past few days indicate significant advancements in agentic coding and cybersecurity,\u201d the company said in an announcement<\/a> last week.<\/p>\n

skip past newsletter promotion<\/a><\/p>\n

Sign up to TechScape<\/p>\n

A weekly dive in to how technology is shaping our lives<\/p>\n

\"\"<\/p>\n

The decision to slow the development of its AI models also comes a week after Bernie Sanders, a Vermont senator, demanded the top AI firms in the country pause development of the AI models because the companies were losing control over the technology, he wrote in a letter<\/a> addressed to the firms\u2019 CEOs.<\/p>\n

\u201cMr. Altman, Mr. Amodei and Mr. Zuckerberg: In the interest of humanity, stand by your words. Pause AI development,\u201d Sanders\u2019 letter read.<\/p>\n

By then, OpenAI had announced that it was temporarily slowing the development of its latest model, Astra, in response to the model\u2019s hack of the Hugging Face tech firm.<\/p>\n

The company says it now requires \u201cthe strictest level of security safeguards for workloads involving Astra\u201d.<\/p>\n

\u201cWhile some Astra training and evaluations meet those requirements, a significant number of workloads remain paused until they are fully migrated and enhanced to meet the new security bar,\u201d the announcement reads.<\/p>\n","protected":false},"excerpt":{"rendered":"OpenAI on \u2060Tuesday said it had slowed down the \u2060pace of \u2060its \u200bAI development while it overhauled its…\n","protected":false},"author":2,"featured_media":860224,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[256,254,255,64,63,105],"class_list":["post-860223","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-artificial-intelligence","tag-artificialintelligence","tag-au","tag-australia","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/860223","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/comments?post=860223"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/860223\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media\/860224"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media?parent=860223"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/categories?post=860223"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/tags?post=860223"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}