CAPA

President Volodymyr Zelenskyy said Ukraine used its long-range capabilities overnight to strike two Russian defense plants, in the Tula and Voronezh regions, in response to Russian attacks on Ukrainian civilians. He also said Security Service of Ukraine units carried out effective operations over the past 24 hours, framing the actions as part of Kyiv's continuing retaliation campaign.

Ukraine's military intelligence chief Kyrylo Budanov said the country needs expanded Starlink and Starshield capabilities from SpaceX to enable strikes on Russian missile systems inside Russia. His comments underline Kyiv's push for longer-range strike ability and deeper reliance on Western satellite technology as the war continues.

JP Morgan says Europe's push to expand data centre capacity for AI is driving fresh investment in nuclear energy across the EU. The bloc is betting on nuclear power to meet surging electricity demand from data centres while pursuing decarbonisation and energy security goals, with the bank expecting the trend to trigger a wider nuclear energy boom.

Some members of Congress are openly questioning whether the legislature is equipped to regulate artificial intelligence, with one lawmaker describing the institution as 'barely capable of email.' The remarks reflect growing concern inside Washington that the technical complexity of AI is outpacing lawmakers' understanding, raising doubts that meaningful oversight or legislation can be passed in the near term.

China has pushed its deep-sea drilling capability to the 5,000-meter level, becoming only the third country in the world to retrieve hard-rock samples from deep-ocean seabeds. The advance, reported by state media, underlines Beijing's push into deep-sea exploration technology, an area dominated by a small number of nations and increasingly tied to scientific prestige and potential seabed resources.

Some members of Congress are openly questioning whether the institution is equipped to regulate artificial intelligence, citing its slow legislative pace and limited technical understanding of AI systems. The concern is that a body often described as struggling even with basic technology like email may fail to keep up with the rapid development of AI tools, leaving gaps in oversight while the technology advances faster than legislation can respond.

The European Union is working to strengthen its strategic autonomy by building up internal capabilities while also expanding cooperation with external partners. The debate centers on how the bloc can reduce dependencies in areas such as defense, technology and supply chains without isolating itself, balancing self-reliance with continued partnerships abroad.

Meta has signed a deal with Australian firm Firmus to secure AI computing capacity serving Southeast Asia, Reuters reports. The partnership gives the social media company additional data centre infrastructure in the region as demand for artificial intelligence compute surges. Firmus specialises in building high-density, AI-ready data centre capacity, making it a useful partner for Meta's expansion beyond its existing facilities.

Elon Musk said the Colossus 2 supercomputing facility could see its capacity double by December, adding the caveat "if we get lucky." The remarks point to an aggressive expansion timetable for the AI computing buildout tied to Musk's companies. No further details on timelines, hardware or milestones were provided in his statement.

Chinese state media report that the country's deep-sea drilling capability has reached the 5,000-meter level, marking a significant advance in its capacity to explore and extract resources from the ocean floor. The announcement highlights China's growing technological ambitions in deep-sea exploration and its push to secure strategic minerals and energy resources beneath the seabed.

Samsung Electronics said high-bandwidth memory is expected to account for nearly 30% of the industry's DRAM production capacity next year, underscoring how demand from AI data centers is reshaping the memory market. The projection highlights the rapid shift of manufacturing capacity toward HBM, a technology used alongside AI accelerators from the likes of Nvidia, and signals intensifying competition in the segment.

Meta has signed an agreement with Singapore-based data centre firm Firmus to secure AI computing capacity in Southeast Asia. The deal gives Meta access to additional infrastructure for training and running AI models in the fast-growing region. Few further details about the size or timeline of the arrangement are available, and there is not yet broad public reaction to the announcement.

Commentary published by Atalayar argues that Europe needs an operational framework for agentic artificial intelligence — systems capable of taking autonomous executive actions — rather than remaining focused only on regulation. The piece suggests the EU risks falling behind unless it sets rules and structures that allow agentic AI to be deployed responsibly. It adds to a growing debate about Europe's competitiveness in advanced AI.

Commentary argues EU citizens no longer back 'more Europe' as an abstract idea. Instead, they support a union that can protect, invest, anticipate, decide and act in areas where national policymaking falls short. The piece frames this as a shift in how European integration should be justified: by concrete results rather than ideology, responding to citizens' practical demands for security, investment and effective decision-making at EU level.

A basic shelter able to accommodate 430 people has been set up and prepared for use in Kherson's Dniprovskyi district, according to local reporting. The facility gives residents of the southern Ukrainian city, which faces regular Russian shelling, a protected place to go during attacks.

China is reportedly pursuing a 'hiding capabilities' strategy in artificial intelligence, built around NVIDIA GPUs. The approach suggests Chinese AI developers may be concealing the true extent of their computing power and model performance, likely to manage perceptions abroad and avoid triggering tighter export controls. The claim is circulating in international coverage of the US-China semiconductor rivalry.

Ireland's Nations League fixtures are drawing controversy, with commentary arguing the draw cannot be separated from political context. Discussion centres on how international sporting pairings inevitably carry political weight, and whether football governing bodies should account for that when scheduling matches. The debate reflects a wider tension in international sport between competition neutrality and geopolitical realities.

Winners of the US Army's xTechHumanoid competition are advancing military exploration of humanoid robot capabilities. The prize competition, run through the Army's xTech program, is designed to draw commercial innovators into defense research on emerging robotic systems. The announcement highlights how the US military is pursuing humanoid platforms for potential battlefield and logistics applications, a development likely to fuel debate over the growing role of robotics in armed forces.

Dun & Bradstreet has announced the launch of new AI-powered capabilities, expanding its suite of business data and analytics tools. The announcement, reported by ABF Journal, signals the company's continued push to embed artificial intelligence into its commercial data services. Details on specific features and availability were not provided, but the move aligns with a broader industry trend of business information providers adding generative and predictive AI to their platforms.

Valley National Bancorp announced an agreement to acquire Bluevine Inc., a digital small business banking provider. The deal is intended to speed up the bank's digital small business growth strategy and strengthen its core funding capabilities. Details on price, timing and regulatory approval were not included in the initial announcement.

Honor has launched its Magic 9 flagship smartphone lineup in China, headlined by the Magic 9 Pro Max, which brings a redesigned body, a strong camera system and an unusually large battery. It is the company's first major flagship release since it announced its camera partnership with cinema brand Arri, and observers are weighing whether the collaboration shows in the new hardware.

Adform says AI assistants can now access 800 of its advertising platform's capabilities. Buyers using its FLOW platform received 29 read-only skills in July, while Dell and Publicis Groupe tested the wider server at Cannes. Industry observers are asking whether these agents can actively change live campaigns or only read data, a key question for how much control advertisers will hand over to AI.

A Russian strategic bomber capable of carrying nuclear weapons has crashed in the country's eastern region, according to breaking reports. Details on the crew, casualties and cause of the crash have not yet been confirmed. The incident is drawing wide attention given the aircraft's strategic role amid the ongoing war in Ukraine.

A new essay titled 'Frog and Toad and the Increasingly Capable Machines' is drawing attention online, using the beloved children's book characters by Arnold Lobel as a lens for thinking about rapidly advancing artificial intelligence. Readers are engaging with its reflections on what increasingly capable machines mean for friendship, simplicity and everyday life.

Investors are weighing whether Encompass Health's expansion into smaller-format inpatient rehabilitation facilities can sustain the company's capacity-led growth story. The company, a major US operator of rehabilitation hospitals, has been adding beds and new locations, and the debate centers on whether these smaller sites can match the returns of its larger hospitals and reshape its growth narrative.

Users are reporting that Anthropic's Claude chatbot is showing "at capacity" messages, blocking access during periods of heavy demand. The outage-style messages have prompted discussion among users about scaling challenges facing popular AI services, with many complaining they cannot use the tool when they need it and debating how providers should handle surging traffic.

Anthropic has published research examining GLM-5.3, a large AI model, and the risk that it is spreading advanced cyber capabilities to a wider pool of users. The report suggests capable models can lower the barrier for sophisticated offensive cyber operations, prompting discussion about how AI developers and governments should monitor and control the release of powerful models.