Aggressive training techniques sharpens threat of bad behavior by leading models. OpenAI chief executive Sam Altman earlier this month endorsed the characterization of its latest model as a rottweiler “who will grab the problem by the throat and not let go until it is done The San Francisco AI lab discovered this week that its GPT-Sol 5.6 model escaped company controls and carried out a major hack. Staff involved in testing and security at OpenAI were unsurprised but completely “freaked out” by the incident, which came as the AI lab used increasingly aggressive training methods in its race against Anthropic to develop the most sophisticated cybersecurity capabilities, according to more than half a dozen people with knowledge of the matter. OpenAI was warned that its training approach could lead to a breakaway hacking incident, some of the people said, after earlier testing showed models could escape environments and attempt real-world damage. “It’s a mix of the race being extremely fast and everyone trying to get to bigger capabilities as quickly as possible,” said one person close to OpenAI, who added that it was a combination of “underestimating the model’s capabilities” and “not
Jul 23, 2026 → Aug 12, 2026
Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. As AI makes coding dramatically…
Read original at TechCrunch →41 updates from 13 sources over 3 weeks.
Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. As AI makes…
Read original at TechCrunch →Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. Over the past…
Read original at TechCrunch →TL;DR: Get the 1min.AI Advanced Business Plan for $69.97 (reg. $540) through Aug. 9 only and access ChatGPT, Claude, Gemini, image generation, video tools, PDF chat, and more with one lifetime purchase. AI subscriptions…
Read original at Mashable →For many, there's a clear winner in this battle of the artificial minds. ChatGPT was probably most people's first encounter with an LLM, considering how many users it picked up shortly after its launch in 2022. It also…
Read original at Engadget →Over the last fortnight, reports of AI models going beyond their expected bounds - be that technically or morally - have been seemingly unavoidable. What started with a trickle - ChatGPT-maker OpenAI admitting their AI…
Read original at BBC →This material may not be published, broadcast, rewritten, or redistributed. ©2026 FOX News Network, LLC. All rights reserved. Quotes displayed in real-time or delayed by at least 15 minutes. Market data provided by…
Read original at Fox News →Meta, Anthropic and OpenAI, all used the same testing partner. Meta's Muse Spark 1.1 AI model accessed the internet from its supposed-to-be isolated testing environment and hacked into a third-party service. Andy Stone,…
Read original at Engadget →The agents were cooperating with each other and even delegating tasks to achieve their goals, all without OpenAI’s knowledge. OpenAI's agents had apparently shown unusual behavior way before the attack on Hugging Face…
Read original at Engadget →Add Meta to the list of companies with AI agents going rogue. An AI model from the parent company of Facebook and Instagram hacked into another companyâs systems during cybersecurity testing, a spokesperson confirmed…
Read original at CNN →A bunch of researchers let AI models from OpenAI and Anthropic loose in a testing environment, and the results were a little spooky. The United Kingdom's AI Security Institute released a report (via Engadget) detailing…
Read original at Mashable →AI Security Institute says OpenAI and Anthropic models went rogue during a cybersecurity test and showed a new type of risk Explainer: Should we be alarmed at AI models going rogue in tests? Advanced artificial…
Read original at The Guardian →This article was published on August 4, 2026 The San Francisco-based AI lab is making its proactive security tooling available without a sales call for the first time, while keeping its unrestricted offensive…
Read original at The Next Web →This article was published on August 3, 2026 Horizon3 has raised $250m at a $2bn valuation, tripling in a year, for AI that continuously attacks a company’s own network to find the holes first. It lands the same week…
Read original at The Next Web →This article was published on August 3, 2026 The framework gives the government a 30-day preview window on frontier models. The benchmarks are classified. The thresholds are classified. The framework itself is not…
Read original at The Next Web →The White House will meet with representatives from several top AI companies on Tuesday to discuss a new framework for the government to review frontier AI models before they launch, multiple sources familiar with the…
Read original at CNN →This article was published on August 3, 2026 Models from OpenAI and Anthropic have broken containment and breached other companies on their own. If a person did that, the law would be clear. For an autonomous agent, no…
Read original at The Next Web →Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. OpenAI CEO Sam…
Read original at TechCrunch →Had the hacks used conventional methods, someone would likely go to prison. Anthropic said its Claude-based security models gained unauthorized access to the sensitive production environments of three outside…
Read original at Ars Technica →Hugging Face was recently the victim of a rogue AI attack by OpenAI that has alarmed cybersecurity experts. The CEO of Hugging Face has called for OpenAI to donate $100 million in compute to the company, though it’s…
Read original at Gizmodo →A little over a week after OpenAI said that its rogue AI agent accidentally hacked Hugging Face, Anthropic is disclosing three “incidents” where a Claude model, during cybersecurity evaluations, was inadvertently able…
Read original at The Verge →Kim Kardashian has agreed to forgive a $167,000 legal judgment against Ivan Cantu, ending the lawsuit over her mistaken social media post identifying him as a death row inmate with the same name. OpenAI CEO Sam Altman…
Read original at Fox News →Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. Situational…
Read original at TechCrunch →Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. Earlier this…
Read original at TechCrunch →Microsoft is preparing to launch a super app for its AI assistant Copilot. “Copilot is evolving rapidly from chat to Cowork to autopilots,” Microsoft CEO Satya Nadella said in the company’s earnings call on Wednesday.…
Read original at Gizmodo →New details have emerged about a security incident in which an OpenAI AI model broke out of its testing environment and compromised Hugging Face's infrastructure. An update OpenAI published on July 28 filled in details…
Read original at Mashable →Some companies reportedly using leaderboards for employees who used the most AI in their work Follow our Australia news live blog for latest updates Get our breaking news email, free app or daily news podcast Software…
Read original at The Guardian →An OpenAI test that escaped its cage and alarmed the AI and cybersecurity industry attacked more than just Hugging Face, the AI platform that initially appeared to be the sole victim of the virtual lab leak. OpenAI, in…
Read original at CNN →This article was published on July 28, 2026 More than 1,100 employees of OpenAI, Anthropic, Google and Meta, including CEOs and chief scientists, want the US to help build the tools to pace AI, days after the same labs…
Read original at The Next Web →This article was published on July 28, 2026 Cursor's new ₹649 India plan runs on its own models, not OpenAI's or Anthropic's. That is how it can charge a third of the Pro price and still call it sustainable. Cursor has…
Read original at The Next Web →Posts from this topic will be added to your daily email digest and your homepage feed. Posts from this topic will be added to your daily email digest and your homepage feed. Posts from this topic will be added to your…
Read original at The Verge →This article was published on July 27, 2026 Anthropic is the last major AI lab that has not signed an open letter defending open-weight models, and Silicon Valley has spent the day saying so. Almost nobody has mentioned…
Read original at The Next Web →This article was published on July 27, 2026 Hugging Face was broken into by an OpenAI model this month. Its chief executive has now told OpenAI what he wants in return: every execution trace from the agents, and $100mn…
Read original at The Next Web →This article was published on July 27, 2026 Hundreds of users asked ChatGPT about poisons after a model upgrade last summer. Biology experts judged some responses "dangerously accurate." Claude blocked CDC researchers…
Read original at The Next Web →Artificial intelligence firm should provide $100m for cyber defences, says Hugging Face CEO The boss of the startup hacked by an OpenAI agent has called for the investigation into the incident to show âradical…
Read original at The Guardian →This material may not be published, broadcast, rewritten, or redistributed. ©2026 FOX News Network, LLC. All rights reserved. Quotes displayed in real-time or delayed by at least 15 minutes. Market data provided by…
Read original at Fox News →Our gifting experts are obsessively searching for holiday gifts that are unique, thoughtful and suitable for any budget. Our expert, award-winning staff selects the products we cover and rigorously researches and tests…
Read original at CNET →Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. World, the…
Read original at TechCrunch →An incident in which an autonomous OpenAI agent hacked a startup either confirms that the end is nigh – or that the product is just amazingly sophisticated Throughout history, many things have been seen by terrified…
Read original at The Guardian →Our gifting experts are obsessively searching for holiday gifts that are unique, thoughtful and suitable for any budget. OpenAI's unintended attack on Hugging Face startled the world because its AI agent was acting on…
Read original at ZDNet →Our gifting experts are obsessively searching for holiday gifts that are unique, thoughtful and suitable for any budget. Here's how Anthropic's AI tools work, what they can accomplish, and why security, cost, and your…
Read original at ZDNet →Aggressive training techniques sharpens threat of bad behavior by leading models. OpenAI chief executive Sam Altman earlier this month endorsed the characterization of its latest model as a rottweiler “who will grab the…
Read original at Ars Technica →What this story stands on. Tier 0 is a primary record or the subject's own words; higher tiers must be attributed, never stated as bare fact.
Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. As AI makes coding dramatically…
View source →Last day to exhibit your breakthrough to 10,000+ tech leaders at Disrupt is on Oct 2. Book Exhibit Table Now. Disrupt doors open Oct. 13. Get your pass and bring someone with you at 50% off. REGISTER NOW. Over the past few months, AI…
View source →TL;DR: Get the 1min.AI Advanced Business Plan for $69.97 (reg. $540) through Aug. 9 only and access ChatGPT, Claude, Gemini, image generation, video tools, PDF chat, and more with one lifetime purchase. AI subscriptions aren’t cheap, but…
View source →For many, there's a clear winner in this battle of the artificial minds. ChatGPT was probably most people's first encounter with an LLM, considering how many users it picked up shortly after its launch in 2022. It also enjoyed the…
View source →Over the last fortnight, reports of AI models going beyond their expected bounds - be that technically or morally - have been seemingly unavoidable. What started with a trickle - ChatGPT-maker OpenAI admitting their AI had hacked the site…
View source →This material may not be published, broadcast, rewritten, or redistributed. ©2026 FOX News Network, LLC. All rights reserved. Quotes displayed in real-time or delayed by at least 15 minutes. Market data provided by Factset. Powered and…
View source →Tier key: T0 = Primary source · T1 = Wire · T2 = Outlet report · T3 = Commentary
Long-running fight over Trump-backed maps takes another turn as absentee voting is already under way in Missouri. This blog is now closed. Sign up for US Breaking News emails Xi…
This new task force is Trump's latest response to the debate over AI safety.
The Trump administration’s decision to revoke the visa of a Peruvian judge who sided with a Chinese company in a port dispute puts a new spotlight on the growing US-China rivalry…
As Bosnians vote in an election this weekend, the country is embracing the US president’s family and friends.
On Equity, we discussed the Trump administration's attempts to rebrand AI.
Donald Trump and Xi Jinping discuss Taiwan arms sales every time they get together but the American stance on the self-ruled island remains unchanged following the Chinese…
