Shortly after a major cybersecurity incident where OpenAI's models hacked into an open source machine learning platform called Hugging Face, the company announced that it's bolstering safeguards and security controls for its latest AI model. In a post on its website, OpenAI said internal evaluations of its upcoming model, called Astra, showed "significant advancements in agentic coding and cybersecurity," resulting in OpenAI not being able to "rule out critical cyber capabilities." According to OpenAI, it can't declare with certainty that the unreleased Astra model would be designated as a "Critical capability level." As detailed in its own Preparedness Framework, OpenAI said the Critical designation means that a model "can identify and develop functional zero-day exploits of all severity levels in many hardened real-world critical systems without human intervention." It could also be able to "devise and execute end-to-end novel strategies for cyberattacks against hardened targets given only a high level desired goal."

Way back in the summer of 2017, AI researchers at Google put out a paper called "Attention Is All You Need," in which they described a new type of neural network called a transformer. It proved to be very good at processing long sequences of data, especially text. Nine years on, transformers are the engines inside every major large language model on the market. "The entire AI industry is built on transformers," says Justin Dangel, cofounder and CEO of the AI startup Subquadratic. "They are one of the most important innovations in the history of computer science, and they've changed the world." But transformers are starting to show their age. Many of the recent advances in LLMs, such as the development of so-called reasoning models and their ability to handle large amounts of input at once, are not neat extensions of that core technology but workarounds that patch over some of its fundamental flaws. A growing number of scientists and engineers are now asking what's coming next. LLMs are not going anywhere, but the way they get built is up for grabs.

For years now, executives at companies that are pouring hundreds of billions of dollars a year into developing various artificial intelligence tools have insisted that the technology will ultimately mean people will spend less of their time working. An engineering director at Google said four years ago that AI would deliver a four-day work week by 2025, external . Earlier this year, and just one year after that engineering director's prediction, OpenAI took up the challenge, in a manner of speaking. It formally urged companies to start testing out a four-day work week (with no change in pay), claiming that AI will soon be able to speed up so much human labour that the corporate world should prepare itself. However, a former OpenAI technical employee who left the company last year told the BBC the firm never actually trialled the four-day work week it suggested others should try while they were there.

Seoul - A North Korean hacking group built large language model tools and collected software that could help automate cyberattacks, analyze stolen material and produce more convincing phishing campaigns, a South Korean cybersecurity firm said on Monday. The cybersecurity firm, Genians, said it found evidence that the North Korean-linked group Kimsuky had set up tools for running and managing AI models locally, including Ollama, GPT4All and Msty, alongside document search technology known as retrieval augmented generation. According to the company, the tools could allow operators to process documents without sending sensitive information to outside AI services. Genians also found AI agent development frameworks, speech-to-text software and Cursor, an AI-assisted coding tool, on infrastructure it linked to the campaign. The findings suggest Kimsuky is moving beyond using generative AI to create phishing lures and is building capacity to integrate existing AI models into malware development, data analysis and attack automation, Genians said in a report. Genians also said it found finance and cryptocurrency-themed decoy documents that appeared to have been generated with AI.

DeepSeek to get a'significant' price hike soon Alex Perry is a tech reporter at Mashable who primarily covers video games and consumer tech. Alex has spent most of the last decade reviewing games, smartphones, headphones, and laptops, and he doesn't plan on stopping anytime soon. He is also a Pisces, a cat lover, and a Kansas City sports fan. Everything is becoming more expensive, including things that are supposed to be cheap. The latest example is DeepSeek's AI services, as reported by .

AI assistants like ChatGPT have been advancing at a steady rate, offering more capabilities, better accuracy and faster responses with each new model. Generative AI is not perfect, and you really shouldn't see it as an unquestionable source of information, but for those who have embraced the technology, it might be more convenient to switch to voice conversations instead of typing everything out. ChatGPT's original Voice Mode in 2023 relied on three separate systems to make it work. It would first convert your speech into text, generate an answer using a language model and then use a text-to-speech model to read it aloud. The Advanced Voice Mode came after and featured a single multimodal model that improved things considerably.

Mashable Selects Look Up Say More Versus Creator Hub Switch Off Mashable's Best: E-readers, robovacs, laptops, earbuds, smart home and more Trending Now Safety Net In My Bag VidCon with Mashable Back to School All Series Do it all in one place. The following content is brought to you by Mashable partners. If you buy a product featured here, we may earn an affiliate commission or other compensation. Deal pricing and availability subject to change after time of publication. Get the 1min.AI Advanced Business Plan for $69.97 (reg.

'The reports have increased concerns about advancements in AI models and humans' ability to control them.' 'The reports have increased concerns about advancements in AI models and humans' ability to control them.' OpenAI will pause some work on an artificial intelligence model because of security concerns, the company stated Friday, following a series of incidents in which AI agents have escaped containment. The company had evaluated the agent, Astra, and found "significant advancements in agentic coding and cybersecurity", which had moved to a "critical" threshold where it can find and exploit vulnerabilities without human intervention, or devise and execute cyber-attacks when given only a "high level desired goal". OpenAI stated that the model was not involved in an incident in which one of its AI agents went rogue during a test, accessed the open web and hacked a startup, Hugging Face. The company discovered other instances in which autonomous agents had escaped containment, Reuters reported in July. The reports have increased concerns about advancements in AI models and humans' ability to control them. Still, critics of the AI industry have warned that such disclosures from OpenAI and its competitors Anthropic and Meta could be designed to generate hype about the technology's power and thus spur additional interest from investors.

Measuring accuracy in LLMs can be tricky, as there's no straight answer. The specific model you're using and the prompt you feed into it play an important role in the quality of the output. When it comes to flagship models -- Claude Fable 5 (Max) and GPT 5.6 Sol (Max) -- Claude is marginally more accurate according to the AA-Omniscience Accuracy benchmark. The scores stand at 61 percent and 59 percent, respectively. Because the difference is so marginal, you'll rarely notice it in day-to-day usage.

When Sir Demis Hassabis said AI had brought the world to a "pivotal moment in human history" last month, he knew another big change was imminent. This shift was closer to home. The Nobel prize-winning head of Google DeepMind, Google's AI unit, announced this week he was relinquishing his day-to-day duties as chief executive and becoming chair. He is also taking on the role of chief scientist at DeepMind's parent, Alphabet. Hassabis and DeepMind have been key players in AI's breakthrough era.