Critical infrastructure

IAEA launches initiative to protect AI in nuclear facilities

The International Atomic Energy Agency (IAEA) has launched a new research project to strengthen computer security for AI in the nuclear sector. The initiative aims to support safe adoption of AI technologies in nuclear facilities, including small modular reactors and other applications.

AI and machine learning systems are increasingly used in the nuclear industry to improve operational efficiency and enhance security measures, such as threat detection. These technologies bring risks like data manipulation or misuse, requiring strong cybersecurity and careful oversight.

The Coordinated Research Project (CRP) on Enhancing Computer Security of Artificial Intelligence Applications for Nuclear Technologies will develop methodologies to identify vulnerabilities, implement protection mechanisms, and create AI-enabled security assessment tools.

Training frameworks will also be established to develop human resources capable of managing AI securely in nuclear environments.

Research organisations from all IAEA member states are invited to join the CRP. Proposals must be submitted by 30 November 2025, with participation encouraged for women and young researchers. The IAEA offers further details through its CRP contact page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot! 

China leads the global generative AI adoption with 515 million users

In China, the use of generative AI has expanded unprecedentedly, reaching 515 million users in the first half of 2025.

The figure, released by the China Internet Network Information Centre, shows more than double the number recorded in December and represents an adoption rate of 36.5 per cent.

Such growth is driven by strong digital infrastructure and the state’s determination to make AI a central tool of national development.

The country’s ‘AI Plus’ strategy aims to integrate AI across all sectors of society and the economy. The majority of users rely on domestic platforms such as DeepSeek, Alibaba Cloud’s Qwen and ByteDance’s Doubao, as access to leading Western models remains restricted.

Young and well-educated citizens dominate the user base, underlining the government’s success in promoting AI literacy among key demographics.

Microsoft’s recent research confirms that China has the world’s largest AI market, surpassing the US in total users. While the US adoption has remained steady, China’s domestic ecosystem continues to accelerate, fuelled by policy support and public enthusiasm for generative tools.

China also leads the world in AI-related intellectual property, with over 1.5 million patent applications accounting for nearly 39 per cent of the global total.

The rapid adoption of home-grown AI technologies reflects a strategic drive for technological self-reliance and positions China at the forefront of global digital transformation.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Google Cloud and NVIDIA join forces to accelerate enterprise AI and industrial digitalisation

NVIDIA and Google Cloud are expanding their collaboration to bring advanced AI computing to a wider range of enterprise workloads.

The new Google Cloud G4 virtual machines, powered by NVIDIA RTX PRO 6000 Blackwell GPUs, are now generally available, combining high-performance computing with scalability for AI, design, and industrial applications.

An announcement that also makes NVIDIA Omniverse and Isaac Sim available on the Google Cloud Marketplace, offering enterprises new tools for digital twin development, robotics simulation, and AI-driven industrial operations.

These integrations enable customers to build realistic virtual environments, train intelligent systems, and streamline design processes.

Powered by the Blackwell architecture, the RTX PRO 6000 GPUs support next-generation AI inference and advanced graphics capabilities. Enterprises can use them to accelerate complex workloads ranging from generative and agentic AI to high-fidelity simulations.

The partnership strengthens Google Cloud’s AI infrastructure and cements NVIDIA’s role as the leading provider of end-to-end computing for enterprise transformation.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

China’s Unitree reveals next-generation humanoid ahead of major IPO

Unitree Robotics has unveiled its most lifelike humanoid robot to date, marking a bold step forward in the country’s rapidly advancing robotics industry.

The new H2 humanoid model, showcased in a short social media video, demonstrated remarkable agility and expressiveness, performing intricate dance moves with striking humanlike grace.

The 180cm-tall, 70kg robot features a silver face with defined eyes, lips and nose, alongside the tagline ‘Destiny Awakening – born to serve everyone safely and friendly’.

A model that represents the company’s growing ambition as it prepares for a mainland listing valued at around US$7 billion.

Unitree’s progress underscores the growing strength of China in humanoid robotics, a field increasingly dominated by domestic innovation and manufacturing capabilities.

As global competition intensifies, the company aims to position itself at the forefront of human-robot interaction and industrial automation.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Startup raises $9m to orchestrate Gulf digital infrastructure

Bilal Abu-Ghazaleh has launched 1001 AI, a London–Dubai startup building an AI-native operating system for critical MENA industries. The two-month-old firm raised $9m seed from CIV, General Catalyst and Lux Capital, with angels including Chris Ré, Amjad Masad and Amira Sajwani.

Target sectors include airports, ports, construction, and oil and gas, where 1001 AI sees billions in avoidable inefficiencies. Its engine ingests live operational data, models workflows and issues real-time directives, rerouting vehicles, reassigning crews and adjusting plans autonomously.

Abu-Ghazaleh brings scale-up experience from Hive AI and Scale AI, where he led GenAI operations and contributor networks. 1001 borrows a consulting-style rollout: embed with clients, co-develop the model, then standardise reusable patterns across similar operational flows.

Investors argue the Gulf is an ideal test bed given sovereign-backed AI ambitions and under-digitised, mission-critical infrastructure. Deena Shakir of Lux says the region is ripe for AI that optimises physical operations at scale, from flight turnarounds to cargo moves.

First deployments are slated for construction by year-end, with aviation and logistics to follow. The funding supports early pilots and hiring across engineering, operations and go-to-market, as 1001 aims to become the Gulf’s orchestration layer before expanding globally.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

AWS outage shows the cost of cloud concentration

A single fault can bring down the modern web. During the outage on Monday, 20 October 2025, millions woke to broken apps, games, banking, and tools after database errors at Amazon Web Services rippled outward. When a shared backbone stumbles, the blast radius engulfs everything from chat to commerce.

The outage underscored cloud concentration risk. Roblox, Fortnite, Pokémon Go, Snapchat, and workplace staples like Slack and Monday.com stumbled together because many depend on the same region and data layer. Failover, throttling, and retries help, but simultaneous strain can swamp safeguards.

On Friday, 19 July 2024, a faulty CrowdStrike update crashed Windows machines worldwide, triggering blue screens that grounded flights, delayed surgeries, and froze point-of-sale systems. The fix was simple; recovery wasn’t. Friday patches gained a new cautionary tale.

Earlier shocks foreshadowed today’s scale. In 1997, a Network Solutions glitch briefly hobbled .com and .net. In 2018, malware in Alaska’s Matanuska-Susitna knocked services offline, sending a community of 100,000 back to paper. Each incident showed how mundane errors cascade into civic life.

Resilience now means multi-region designs, cross-cloud failovers, tested runbooks, rate-limit backstops, and graceful read-only modes. Add regulatory stress tests, clear incident comms, and sector drills with hospitals, airlines, and banks. The internet will keep breaking; our job is to make it bend.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

SMEs underinsured as Canada’s cyber landscape shifts

Canada’s cyber insurance market is stabilising, with stronger underwriting, steadier loss trends, and more product choice, the Insurance Bureau of Canada says. But the threat landscape is accelerating as attackers weaponise AI, leaving many small and medium-sized enterprises exposed and underinsured.

Rapid market growth brought painful losses during the ransomware surge: from 2019 to 2023, combined loss ratios averaged about 155%, forcing tighter pricing and coverage. Insurers have recalibrated, yet rising AI-enabled phishing and deepfake impersonations are lifting complexity and potential severity.

Policy is catching up unevenly. Bill C-8 in Canada would revive critical-infrastructure cybersecurity standards, stronger oversight, and baseline rules for risk management and incident reporting. Public–private programmes signal progress but need sustained execution.

SMEs remain the pressure point. Low uptake means minor breaches can cost tens or hundreds of thousands, while severe incidents can be fatal. Underinsurance shifts shock to the wider economy, challenging insurers to balance affordability with long-term viability.

The Bureau urges practical resilience: clearer governance, employee training, incident playbooks, and fit-for-purpose cover. Education campaigns and free guidance aim to demystify coverage, boost readiness, and help SMEs recover faster when attacks hit, supporting a more durable digital economy.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

AI-generated images used in jewellery scam

A jeweller in Hove is dealing with daily complaints from customers of a similarly named but fraudulent business. Stevie Holmes runs Scarlett Jewellery but keeps receiving complaints from customers who confused it with the AI-driven Scarlett Jewels website.

Many reported receiving poor-quality goods or nothing at all.

Holmes said the mix-ups have kept her occupied for at least an hour a day since July. Without clarification, people could post negative comments about her genuine business on social media, potentially damaging its reputation.

Scarlett Jewels is run by Denimtex Limited with an address in Hong Kong, though its website claims a personal story of a retiring designer.

Experts say such scams are increasingly common due to how easy and cheap it is to create AI images. Professor Ana Canhoto from the University of Sussex noted AI-generated product photos often appear too perfect or flawed, while fake reviews and claims of scarcity are typical tactics to mislead buyers.

Trustpilot ratings for Scarlett Jewels are mostly one star, with customers describing items as ‘tat’ or ‘poor quality’.

Authorities are taking action, with the Advertising Standards Authority banning similar ads and Facebook restricting Scarlett Jewels from creating new adverts. Buyers are advised to spot off AI images, large discounts, and genuine reviews to avoid falling for scams.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Bitcoin wallet vulnerability exposes thousands of private keys

A flaw in the widely used Libbitcoin Explorer (bx) 3.x series has exposed over 120,000 Bitcoin private keys, according to crypto wallet provider OneKey. The flaw arose from a weak random number generator that used system time, making wallet keys predictable.

Attackers aware of wallet creation times could reconstruct private keys and access funds.

Several wallets were affected, including versions of Trust Wallet Extension and Trust Wallet Core prior to patched releases. Researchers said the Mersenne Twister-32’s limited seed space let hackers automate attacks and recreate private keys, possibly causing past fund losses like the ‘Milk Sad’ cases.

OneKey confirmed its own wallets remain secure, using cryptographically strong random number generation and hardware Secure Elements certified to global security standards.

OneKey also examined its software wallets, ensuring that desktop, browser, Android, and iOS versions rely on secure system-level entropy sources. The firm urged long-term crypto holders to use hardware wallets and avoid importing software-generated mnemonics to reduce risk.

The company emphasised that wallet security depends on the integrity of the device and operating environment.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Meta champions open hardware to power the next generation of AI data centres

The US tech giant, Meta, believes open hardware will define the future of AI data centre infrastructure. Speaking at the Open Compute Project Global Summit, the company outlined a series of innovations designed to make large-scale AI systems more efficient, sustainable, and collaborative.

Meta, one of the OCP’s founding members, said open source hardware remains essential to scaling the physical infrastructure required for the next generation of AI.

During the summit, Meta joined industry peers in supporting OCP’s Open Data Center Initiative, which calls for shared standards in power, cooling, and mechanical design.

The company also unveiled a new generation of network fabrics for AI training clusters, integrating NVIDIA’s Spectrum Ethernet to enable greater flexibility and performance.

As part of the effort, Meta became an initiating member of Ethernet for Scale-Up Networking, aiming to strengthen connectivity across increasingly complex AI systems.

Meta further introduced the Open Rack Wide (ORW) form factor, an open source data rack standard optimised for the power and cooling demands of modern AI.

Built on ORW specifications, AMD’s new Helios rack was presented as the most advanced AI rack yet, embodying the shift toward interoperable and standardised infrastructure.

Meta also showcased new AI hardware platforms built to improve performance and serviceability for large-scale generative AI workloads.

Sustainability remains central to Meta’s strategy. The company presented ‘Design for Sustainability’, a framework to reduce hardware emissions through modularity, reuse, and extended lifecycles.

It also shared how its Llama AI models help track emissions across millions of components. Meta said it will continue to

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!