Tech Policy EU
The AI Act obligations for providers of general-purpose AI (GPAI) models have entered into application across the EU. To assist providers, the Commission published guidelines clarifying who must comply with the obligations. GPAI models are defined as those trained with over 10^23 FLOP and capable of generating language. Additionally, the Commission published a template to help providers in summarising the data used to train their models.
Elon Musk’s xAI on Thursday said it will sign a chapter on safety and security from the European Union’s code of practice, which aims to help companies comply with the bloc’s landmark artificial intelligence rules. The EU’s code has three chapters – transparency, copyright and safety and security. While the guidance on transparency and copyright will apply to all general-purpose AI providers, the chapters on safety and security target providers of the most advanced models.
Big Tech’s acquihire deals face regulatory scrutiny, outgoing EU antitrust official says
Big Tech deals to acquire skills rather than major companies may soon come under the regulatory scrutiny they previously avoided, the outgoing head of the European Commission’s antitrust unit said. Acquihires, in which Big Tech hires start-ups’ founders and senior managers rather than acquire the companies, have been viewed by antitrust regulators as an attempt to evade merger rules.
Europe sets its sights on multi-billion-euro gigawatt factories as it plays catch-up on AI
The gigafactories could add 15% to Europe’s total computing capacity — a sizeable boost, even when compared to the U.S. which currently owns around a third of global capacity according to UBS data. The bloc has long been a laggard behind the U.S. and China in the race to scale up artificial intelligence. With 27 members in the union, the region is slower to act when it comes to agreeing new legislation. Higher energy costs, permitting delays and a grid in dire need of modernization can also hamper developments. Henna Virkkunen, the European Commission’s executive vice president for tech sovereignty, told CNBC that the bloc’s goal is to bring together high quality data sets, computing capacity and researchers, all in one place.
Swedish PM under fire for using AI in role
The Swedish prime minister, Ulf Kristersson, has come under fire after admitting that he regularly consults AI tools for a second opinion in his role running the country. “You have to be very careful,” Simone Fischer-Hübner, a computer science researcher at Karlstad University, told Aftonbladet, warning against using ChatGPT to work with sensitive information.
Europe’s tech sovereignty watch (Market research report from Proton)
Proton released a comprehensive market report indicating that over 74% of all publicly listed European companies depend on US-based tech services, like Google and Microsoft. Proton analyzed business email domains across Europe to show how many publicly listed companies are reliant on US email and email security services. Proton looked into e-mails because it’s often the gateway to a company’s tech stack. When a company chooses an email service, it often uses the entire suite.
Tech Policy US
Trump administration dials up attack on EU’s digital rulebook
US Secretary of State Marco Rubio has instructed US diplomats to work to undermine the EU’s online moderation rules, according to a diplomatic cable seen by Reuters. The State Department cable, sent by US Secretary of State Marco Rubio on 4 August, instructs the country’s diplomats to try “to repeal and/or amend” the EU’s Digital Services Act (DSA), Reuters reported on Thursday. The US government has also just announced a 100% global tariff on semiconductors – days after the EU said it had agreed a general 15% levy with the US, including for EU-made chips. Trump announced the 100% tariff on imports of semiconductors on 6 August, exempting only companies that manufacture chips in the US or commit to doing so. However the EU still believes its agreement with the US stands: Gill stressed that the US committed to a 15% tariff ceiling applying to EU exports on semiconductors “irrespective of other tariffs”.
U.S. approves OpenAI Google Anthropic as federal AI vendors
The U.S. government has selected OpenAI, Google, and Anthropic as approved artificial intelligence (AI) vendors for federal agencies, streamlining the procurement process for AI tools through the General Services Administration (GSA)’s Multiple Award Schedule program. This initiative removes previous bureaucratic hurdles that required lengthy legal and procurement reviews for agencies seeking AI solutions. The three companies, creators of widely used models such as ChatGPT, Gemini, and Claude, have undergone evaluations for security, performance, and bias… While specific contract amounts remain undisclosed, the inclusion of U.S.-based AI providers highlights the administration’s focus on domestic innovation in a competitive global landscape. The Pentagon, which has already contracted with AI firms for defense-related projects, is part of a growing federal consensus that AI is a core strategic asset.
The day Silicon Valley’s oligarchs kissed Trump’s ring
Le Monde launched the first installment of a six-part series examining the rise of tech oligarchs in the U.S. and their growing influence over government, regulation, and democratic institutions. Part one focuses on Silicon Valley’s shifting political alliances, offering a historical perspective on how key industry figures have aligned themselves with power. It sets the stage for a deeper exploration of how tech and politics are becoming increasingly intertwined.
Tech Policy Other Regions
Taiwan investigates 16 Chinese firms for poaching high-tech talent
Taiwan authorities said on Thursday they were investigating 16 Chinese companies for allegedly poaching semiconductor and other high-tech talent, amid growing concerns over technology outflows. More than 300 agents searched 70 locations and questioned 120 people between July 15 and August 6, the Taiwan Investigation Bureau said, in coordinated raids on Chinese firms suspected of illegally recruiting engineers and operating unlicensed offices in Taiwan.
China’s NDRC pushes chip sector consolidation in response to U.S. export controls
China’s government convened a group of chip equipment makers this year to discuss a potential megamerger that would combine different technologies into a single state-backed giant, according to multiple people familiar with the talks. The merger plan, led by the National Development and Reform Commission, is part of a broader policy shift aimed at streamlining China’s semiconductor sector, which Beijing sees as vital for strengthening its domestic chip industry amid continuing US export controls aimed at curbing the country’s high-tech ambitions. There have been 26 semiconductor acquisitions announced so far this year, according to data from financial information provider Wind. The most high-profile deal was a merger announced in May between Hygon, which designs central processing units for servers and data centres, and supercomputer maker Sugon.
The Gulf bets big on AI as it seeks the ‘new oil’
The UAE and Saudi Arabia want to host the data centres needed to train powerful AI models. Gulf sovereign funds have poured billions into foreign tech giants in recent years. But now, they are pivoting from being passive investors to more active players. In Saudi Arabia, the Public Investment Fund (PIF) has launched the national AI company Humain which plans to build “AI factories” powered by several hundred thousand Nvidia chips over the next five years. In the UAE, state-owned investment firm Mubadala has backed G42 and MGX (a $100bn AI-focused joint venture) with Microsoft involved as a key technology provider. Earlier this year, Trump’s visit gave Washington a strong lead in the region’s AI race.. As part of its strategic pivot, the UAE has scaled back some China-backed projects and reduced its reliance on Huawei hardware. The emphasis on AI deals during Trump’s visit underscores the growing strategic importance of the technology to US diplomacy. Traditionally, the US-Gulf relationship has centred around oil-for-security. But the dynamic is now shifting toward energy, security – and tech.
Market Trends
NTT Data and Mistral AI team on sustainable, private AI for regulated enterprises
NTT Data and French AI innovator Mistral AI have unveiled a strategic partnership to deliver secure, sustainable and enterprise-grade private AI solutions, targeting regulated industries such as financial services, insurance, defence, and the public sector. The partnership will focus on three key areas: sustainable and secure co-development of private AI; enhancing IT infrastructure and customer experience through agentic AI, particularly in Europe and Asia Pacific; and joint go-to-market expansion across countries including France, Luxembourg, Spain, Singapore and Australia.
How Google plans to reduce US AI data centre power consumption
The technology giant has partnered with Indiana Michigan Power (I&M, an American Electric Power company) and Tennessee Power Authority in the US. These alliances will involve scaling back power use at Google’s data centres when called upon by the electric utilities to free up space on the grid. First reported by Reuters, this news comes at a significant time for the industry, particularly as the grid continues to experience surging demand. As AI continues to use significant amounts of power, energy and water to run, data centres are now having to reckon with the continued challenge of sustainability and power constraints.
SK Telecom debuts new GPU cluster to boost AI in Korea
SK Telecom has launched one of South Korea’s largest GPU-as-a-Service (GPUaaS) platforms, built on over 1,000 Nvidia B200 GPUs and designed to serve as a cornerstone of the country’s sovereign AI infrastructure… The cluster has been selected to support the Ministry of Science and ICT’s ‘Proprietary AI Foundation Model’ programme, a national initiative to advance Korea’s AI competitiveness and strengthen the local ecosystem. SK Telecom plans to position its Gasan AI Data Centre (AIDC) as a critical node in this effort.
Huawei unveils CloudMatrix 384 to challenge Nvidia’s AI dominance
Huawei’s strategy, however, is notably inward-facing. The system is tightly integrated with Huawei’s proprietary software stack, including MindSpore and CloudMatrix-Infer, and lacks compatibility with the CUDA ecosystem dominant outside China. As such, international uptake is expected to be limited. Nonetheless, the move represents a clear signal: AI infrastructure is now central to digital sovereignty, and Huawei is determined to build its own path even at the cost of power efficiency or global interoperability.
How Delta is using AI for ticket pricing and what it means for air travel
“Surveillance pricing has been shown to utilize extensive personal information obtained through a variety of third party channels, including data about a passenger’s purchase history, web browsing behavior, geolocation, social media activity, biometric data, and financial status,” [senators] wrote. “Former FTC Chair Lina Khan has cautioned against a particularly egregious but conceivable example of an airline using AI to charge a higher fare to a passenger ‘because the company knows that they just had a death in the family and need to fly across the country.’ The opacity surrounding Delta’s new customized pricing model could aggravate these concerns.” A Delta representative told ABC News that Delta is in compliance “with regulations around pricing and disclosures” in its AI technology experimentation. “There is no fare product Delta is testing that is based on individual information,” the representative said in an emailed statement. “Market forces drive the dynamic pricing model that’s been used in the global industry for decades.”
OpenAI introduces Stargate Norway, its first AI data centre initiative in Europe
Stargate Norway partners include Nscale, a leading AI infrastructure provider, and Aker, a legacy leader in energy and industry. The data centre site will be designed and built by Nscale and is expected to be owned as part of a 50/50 joint venture between Nscale and Aker. Touted as one of the most ambitious AI infrastructure investments in Europe to date, Stargate Norway hopes to harness the town of Narvik’s abundant hydropower, low-cost energy, cool climate and strong industrial base. Both Aker and Nscale are working to provide priority access to Norway’s AI ecosystem, OpenAI says, to ensure that local AI start-ups and scientific researchers can benefit from the additional compute capacity that will be on offer.
Tesla taps Samsung for $16.5bn AI chip megadeal as Texas fab ramps up
The decade-long deal, first revealed in South Korean regulatory filings last week, sees Samsung’s foundry division take on fabrication duties for Tesla’s sixth-generation AI chips, branded internally as “AI6”. The chips, custom-designed by Tesla engineers, will power future autonomous driving systems, robotics platforms, and Tesla’s Dojo-class training clusters. The deal marks a critical breakthrough for Samsung Foundry, which has struggled in recent years to match the yields and scale of Taiwan’s TSMC. Winning Tesla, a notoriously exacting client, as a customer lends Samsung renewed credibility as it competes for next-gen AI workloads. The move also signals Tesla’s deepening intent to vertically integrate its AI stack, from chip design to software training. and reduce dependence on third-party GPU suppliers such as Nvidia.
Inside the relentless race for AI capacity
The Financial Times published a visual story about the explosive growth of AI infrastructure and how AI data centers are multiplying at a historic pace to meet the demand for compute power. Microsoft, Alphabet, Amazon and Meta plan to increase their capital expenditures to more than $300bn in 2025. IT consultancy Gartner reckons a total of $475bn will be spent on data centres this year, up 42 per cent on 2024. Some believe even that will not be enough to keep up with global demand for AI. McKinsey predicted in April that $5.2tn of investment in data centres would be required by 2030. Yet building the facilities required to pursue this next frontier while also supporting billions of people using sophisticated AI models is far more complex, costly and energy-guzzling than the power grids, telecoms networks or even cloud computing systems that came before them.
Geopolitics
Israel relying on Microsoft cloud for expansive surveillance of Palestinians
The Guardian reported on an investigation revealing that the Israeli military undertook an ambitious project to store a giant trove of Palestinians’ phone calls on Microsoft’s servers in Europe. The cloud-based system – which first became operational in 2022 – enables Unit 8200 (Israel’s military surveillance agency) to store a giant trove of calls daily for extended periods of time. Microsoft claims its chief executive Satya Nadella was unaware of what kind of data Unit 8200 planned to store. But a cache of leaked Microsoft documents and interviews with 11 sources from the company and Israeli military intelligence reveals how Azure has been used by Unit 8200 to store this expansive archive of everyday Palestinian communications. According to three Unit 8200 sources, the cloud-based storage platform has facilitated the preparation of deadly airstrikes and has shaped military operations in Gaza and the West Bank.
Trump threatens new sanctions as Putin meets US envoy in Moscow
Russian President Vladimir Putin met US envoy Steve Witkoff in Moscow on Wednesday, August 6, the Kremlin said, two days ahead of a deadline imposed by US President Donald Trump on Moscow to halt its Ukraine offensive or face fresh sanctions. Trump… has given Russia until Friday to make progress toward peace or face new penalties. Moscow has demanded that Ukraine cede more territory and renounce Western support if it wants the fighting to stop. Kyiv is calling for an immediate ceasefire, and Zelensky last week urged his allies to push for “regime change” in Moscow. Trump has increasingly voiced frustration with Putin in recent weeks over Russia’s unrelenting offensive. Russia fired a record number of long-range drones at Ukraine in July, Agence France-Presse analysis of data from Kyiv’s air force showed. The White House has not outlined specific actions it would take against Russia, but Trump has previously threatened to impose “secondary tariffs” targeting Russia’s key trade partners.
Discover more from Aldeota Global
Subscribe to get the latest posts sent to your email.
Aldeota Global 