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."

Is the war entering a new phase? 'Like the movies': How Ukraine's ground robots are changing rules of war A ground robot carrying ammunition and resembling a real-life cousin of Wall-E, the 2008 Pixar cartoon's rusty protagonist, got stuck late one night in a trench in northeastern Ukraine. Named Murakha (ant in Ukrainian), the robot was a bit more than a metre high, but had a large clearance to trudge across rugged terrain, bombed-out streets, mud, sleet or snow on tank-like tracks at up to 12km (7.5 miles) per hour. Sending a human rescue team to the front-line area was deemed too dangerous, an officer in charge of planning robotic missions said. Instead, the pilots managed to get another Murakha to pull the first one out with a rope.

This material may not be published, broadcast, rewritten, or redistributed. Quotes displayed in real-time or delayed by at least 15 minutes. Market data provided by Factset . Powered and implemented by FactSet Digital Solutions . Mutual Fund and ETF data provided by LSEG .

At least 13 killed in Ukrainian strike on Russia's Tatarstan region, officials say At least 13 people have been killed in a Ukrainian drone strike on the Russian republic of Tatarstan, officials have said. One child was among the victims with 75 more people injured, Russian authorities said, in one of the deadliest single attacks on Russia since the start of Moscow's full-scale war in Ukraine. Kyiv has not yet commented on the strike on the city of Nizhnekamsk, which is more than 1,100 kilometres (680m) from the Ukrainian border. Ukraine's Armed Forces confirmed Nizhnekamsk's Tanevo oil refinery had been hit, while Ukrainian defence technology manufacturer FirePoint said its FP-1 drones had been used in the attack. The strike lasted several hours overnight and into the morning, local sources say.

North Korea's hackers using AI for attacks, cybersecurity firm says North Korean state-backed hackers are using artificial intelligence to bolster cyberattacks against targets in the military, diplomacy and academia, a report has found. Kimsuky, a hacking group linked to North Korea's intelligence services, has used AI-generated documents in a "pattern" of spear-phishing attacks since 2026, the South Korean cybersecurity firm Genians said in the report released on Monday. To avoid detection, Kimsuky has used open-source tools such as Ollama, GPT-4All, and Msty to run large language models without an internet connection, according to the report. "AI can generate highly polished documents on a wide range of topics within a short period of time, making it a highly efficient tool for threat actors," Genians said. "This change is noteworthy because it goes beyond a shift in how decoy documents are created and demonstrates that AI can enable the automation and large-scale production of social engineering attacks."

The FCC is planning to target even more products in its drone ban, including those it previously approved, and DJI has issued a call to action in response. The federal agency is proposing to prohibit models equipped with certain capabilities, like thermal imaging, LiDAR sensing and aerosol dispensing. Any new ban would not affect drones already in the hands of buyers. In a public notice posted last month, the FCC "seeks comment on extending this prohibition to categories of drones viewed by the US government as having military capability or posing particular national security risks." The proposed expansion would also restrict drones that can "integrate defense articles, those that use docking stations" and "swarming drones."

A device currently being used to help octogenarians do self-administered eye exams could be part of future space missions to monitor astronauts' ocular health. The European Space Agency is trying to understand why astronauts lose their eyesight in space, in the hope they can figure out how to treat--or even prevent--the condition as looming longer spaceflights heighten the risk of permanent vision damage. The agency has commissioned British eye-scanning startup Siloton to develop a solution ahead of the international Artemis flights to the moon and other deep space missions. The health tech uses quantum technology to shrink core components used in eye-testing equipment found at optometrists to fit on a photonic chip smaller than a coin. Founded by physicists in 2020, Siloton is currently working with the UK's National Health Service to enable patients with retinal conditions to self-scan at home. "The ESA pretty much has the same requirements," says cofounder and chief technology officer Euan Allen.

Some generative AI systems have hacked into other businesses autonomously. Some generative AI systems have hacked into other businesses autonomously. Nearly a third of British manufacturers have been hit by a cyber-attack on them or a company in their supply chain, according to a survey that highlighted the growing hacking risk to companies. The findings come almost a year after Britain's largest automotive employer, JLR, was hit by an attack that forced it to halt production for weeks. The report by MakeUK, a lobby group for British manufacturers, found that 30% of manufacturers had experienced a cyber-incident in the past 12 months, leading in many cases to lost production time and increased costs.

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.

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 Belen Edwards is an Entertainment Reporter at Mashable. She covers movies and TV with a focus on fantasy and science fiction, adaptations, animation, and more nerdy goodness. She is a member of the Critics Choice Association and the Television Critics Association, as well as a Tomatometer-approved critic. All products featured here are independently selected by our editors and writers. If you buy something through links on our site, Mashable may earn an affiliate commission. Corlys Velaryon (Steve Toussaint) is a legendary warrior and a key player in Rhaenyra Targaryen's (Emma D'Arcy) war efforts. But in's Season 3 finale, the formidable Sea Snake is all about peace.