Home Blog Page 8

Oceans and space to host next wave of data centres due to AI workloads

  • Floating and underwater data centres provide scalable, efficient solutions by using seawater for cooling and by accessing renewable energy near coastlines.
  • Space-based data centres extend these innovations by leveraging the cold of space and abundant solar energy for high-performance computing.

Surging demand from artificial intelligence, cloud services, and digital applications is pushing traditional, land-based data centres to their limits, prompting exploration of oceanic and space-based alternatives, according to GlobalData’s Strategic Intelligence report.

GlobalData says ocean and space environments offer abundant natural cooling, renewable energy access, and flexible deployment models that could ease power and land constraints.

“Floating and underwater data centres provide scalable, efficient solutions by using seawater for cooling and by accessing renewable energy near coastlines,” said Martina Raveni, Strategic Intelligence Analyst at GlobalData.

She added that modular designs can reduce reliance on land and potable water, improve latency for coastal populations, and mitigate climate risks that threaten terrestrial sites.

Pilot projects

Pilot projects in North America, Europe, and Asia have demonstrated the viability of ocean-based deployments at megawatt scales with high rack density and potential for long-term operation. In a notable milestone, China in June 2025 launched its first commercial-scale underwater data centre off Shanghai, powered largely by offshore wind.

The project aims to reach 24 MW across two phases, target power usage effectiveness (PUE) below 1.15, and draw about 97 per cent of its electricity from offshore wind, with seawater cooling integrated into modular submerged units.

Raveni said space-based data centres extend these innovations by leveraging the cold of space and abundant solar energy for high-performance computing. Falling launch costs, reusable rockets, in-orbit networking advances, and optimized orbital choices are improving feasibility. Such platforms could provide low-latency processing and storage for satellites and spacecraft while complementing terrestrial networks.

While the use of ocean-based centres remains limited, GlobalData expects their footprint to grow over the next decade as technologies mature and benefits are validated. Space-based data centres, including orbital and lunar concepts, remain largely experimental, but the firm sees expanding opportunities for partnerships, funding, and research as the ecosystem develops.

France’s ANSSI and UK warn of surging smartphone threats

  • Calls for tighter smartphone security as devices become high‑value targets for cybercriminals, state‑sponsored groups, and private sector offensive actors
  • Report outlines how attackers exploit vulnerabilities in cellular networks, Wi‑Fi, Bluetooth, and NFC to intercept communications, track users, and remotely deploy malware.

France’s National Cybersecurity Agency (ANSSI), in collaboration with the United Kingdom, has issued a joint warning that smartphones have become prime targets for cybercriminals, state-backed groups, and private sector offensive actors, calling for stronger protections across consumer and enterprise use.

The agencies’ report outlines how attackers exploit vulnerabilities in cellular networks, Wi‑Fi, Bluetooth, and NFC to intercept communications, track users, and remotely deploy malware. It flags a growing wave of zero‑click exploits—attacks requiring no user interaction—that often leaves little forensic evidence and complicates detection and response.

Key findings

  • Network exploitation: Threat actors intercept communications, geolocate users, or inject malware via weaknesses in cellular, Wi‑Fi, Bluetooth, and NFC.
  • Zero‑click rise: Exploits requiring no user interaction are increasingly used to compromise targets covertly, complicating detection and incident response.
  • OS/app vectors: Vulnerabilities in operating systems and third‑party apps enable sensitive data collection, which can later fuel targeted phishing and long‑term access.

ANSSI said it has handled multiple incidents in recent years involving compromised mobile devices, citing both unsafe user behaviour and targeted spyware as causes. The report warns that operating systems and mobile apps remain common intrusion vectors for data theft that can later fuel phishing campaigns or provide persistent, covert access to organisational networks.

Recommended countermeasures include disabling wireless features (Wi‑Fi, Bluetooth, NFC) when not needed, avoiding public networks that can be spoofed or monitored, and maintaining rigorous patching of operating systems and applications. Users are advised to rely on strong, unique passwords, restrict app permissions, and favor authenticator apps over SMS for verification.

The guidance also urges hardening devices with platform-specific defenses—such as Lockdown Mode on iOS and Advanced Protection Mode on Android—to reduce the attack surface. ANSSI further recommends non‑biometric passcodes for device protection; the agency did not detail its rationale, though security practitioners often cite risks of biometric spoofing or coercion.

The France–UK advisory underscores escalating risk to both society and the corporate sector, pressing individuals and organisations to elevate mobile security baselines, tighten access controls, and improve incident readiness as smartphone‑centric threats intensify.

What’s next

  • Organisations should audit mobile fleets, enforce hardening at scale, and reassess high‑risk user groups (executives, admins, field teams). Individuals should adopt least‑privilege app permissions, keep radios off by default, and move to app‑based authentication with regular OS updates.
  • Policymakers may push for stronger default protections and clearer disclosure around mobile vulnerabilities, as smartphone‑centric threat activity continues to climb.

e& extends 5G Standalone lead over du on higher capex

  • Performance gains lift UAE’s network leadership globally.

The UAE’s two mobile operators launched 5G Standalone (SA) in 2023, but adoption is advancing faster at e& than at du.

e&’s SA penetration rose from 20 per cent in Q2 2025 to 24 per cent in Q3 2025, reflecting steady progress in nationwide rollout alongside growing device and SIM compatibility. du’s SA uptake remains slower by comparison.

SA is built on a dedicated 5G core—unlike 5G Non‑Standalone (NSA), which relies on 4G control—unlocking capabilities such as ultra‑reliable low‑latency communications (URLLC), network slicing, and stronger uplink performance. This underpins enterprise SLAs, edge computing, and new consumer experiences, and sets the stage for 5G‑Advanced with AI/ML‑driven, self‑optimising networks.

According to Opensignal’s stats, e& users on 5G SA averaged 403.7Mbps download speeds in Q3 2025, 62 per cent higher than the 249Mbps on NSA.

  • Upload speeds: SA reached 29.8 Mbps, 11 per cent above NSA’s 26.8Mbps. Notably, NSA held an early‑year uplink edge (Q1: 31.1Mbps NSA vs. 28.1Mbps SA), but SA overtook by Q3 as uplink scheduling, carrier aggregation, and radio resource management matured on the 5G core.
  • Stability: SA download speeds stayed above 400Mbps across Q2–Q3, while NSA averaged below 270Mbps, indicating SA’s consistency through the period.

Investment gap

  • e&: AED3 billion in Q3 2025 capex (excluding spectrum and licenses), with capital intensity of 16.3 per cent.
  • du: AED492 million in Q3 2025 capex, down from AED511 million in Q3 2024; capital intensity at 12.7 per cent versus 14.2 per cent a year earlier.
  • Context: Industry experts tie e&’s network performance to sustained investment and progressive upgrades, supporting SA densification and core optimisation.

The UAE ranks among top global markets for 5G SA download performance at about 384Mbps in recent benchmarks—well above its NSA results—underscoring the country’s leadership and e&’s contribution during Q3 2025.

Outlook for operators and enterprises

  • Adoption drivers: Continued SA growth depends on broader SA‑capable device portfolios, SIM provisioning, and indoor coverage. Enterprise demand for slicing, deterministic latency, and uplink‑heavy use cases (video sharing, cloud gaming, industrial telemetry) will be pivotal.
  • Competitive implications: e&’s higher capex and measured performance lead position it to capture more SA traffic and premium service tiers. du may need targeted densification, device partnerships, and differentiated enterprise offers to accelerate SA uptake.
  • Next phase: 5G‑Advanced initiatives—embedding AI/ML across RAN and core—are set to improve energy efficiency, automation, and service agility, potentially widening performance gaps for operators that invest early.

Related Posts:

Alibaba launches Quark AI glasses in China

  • Aims to make smart glasses a commerce-centric interface—extending its “super-app” ecosystem into ambient computing and everyday retail interactions.
  • Alibaba must compete not only with global device leaders but also with Chinese peers scaling quickly through aggressive channel strategies.

Alibaba unveiled its Quark artificial intelligence glasses in China, marking the e-commerce giant’s push into AI wearables as global smart glasses shipments accelerate.

Priced from 1,899 yuan (about $268), the Quark glasses are powered by Alibaba’s Qwen AI model and app and are designed to resemble regular eyewear with a black plastic frame, contrasting with bulkier headsets from rivals.

Product and ecosystem integration

  • Features: On-the-go translation, instant price recognition, and voice-first assistance.
  • Ecosystem: Deep integration with Alipay and Taobao positions the device as a commerce- and payments-aware assistant, potentially enabling in-store price checks, product comparisons, and seamless checkout.
  • Availability: Listed on Tmall, JD.com, and Douyin; Alibaba has not yet disclosed sales figures following Thursday’s launch.

Market context

  • Category growth: Global smart glasses shipments reached 4.07 million units in first half of 2025, up 64.2% year over year, according to IDC. The market is forecast to exceed 40 million units by 2029, with Asia Pacific—particularly China—posting the highest growth rate.
  • China’s position: Chinese vendors shipped over 1 million units in first half of 2025, up 64.2% y/y, capturing 26.6% global share. China surpassed the US with a 35.5% market share in H1, ranking first globally, supported by end-to-end supply chain strengths in optics, sensors, and assembly.
  • Competitive landscape: Meta holds roughly 80% of the VR headset market, while Apple sells Vision Pro and Samsung’s Galaxy XR launched in October with Google’s AI features. In China, Xiaomi and Baidu have introduced similar AI-enabled glasses, intensifying local competition.

Strategic implications

  • Differentiation: By embedding Qwen AI and linking directly to Alipay/Taobao, Alibaba aims to make smart glasses a commerce-centric interface—extending its “super-app” ecosystem into ambient computing and everyday retail interactions.
  • Go-to-market: Domestic dominance plus supply chain leverage could support rapid cost iteration and feature updates, while distribution via major Chinese e-commerce platforms provides immediate reach. Overseas expansion will hinge on localised services, app partnerships, and regulatory alignment.
  • Risks: User adoption will depend on comfort, battery life, optical performance, and privacy assurances. Alibaba must compete not only with global device leaders but also with Chinese peers scaling quickly through aggressive channel strategies.

Outlook IDC’s projections and China’s supply chain advantages suggest smart glasses could become a mainstream human–computer interface by decade’s end, providing new growth vectors for consumer electronics.

Underground AI models like WormGPT and KawaiiGPT resurface

  • Analysts say defenders should expect more polished phishing at scale and quicker prototyping of commodity malware
  • Organisations urged to harden identity controls, email authentication, and script execution policies as these underground tools evolve.

Hackers are increasingly adopting large language models tailored for cyberattacks, with tools such as WormGPT and KawaiiGPT re-emerging on dark-web forums and developer platforms, according to new research.

Marketed as “AI without boundaries,” these models promise faster generation of phishing emails, malicious code, and basic ransomware scripts, and are being sold via subscription tiers and supported by active online communities.

Researchers found that many of these underground models appear to be repackaged versions of larger commercial systems but are trained on malware-heavy datasets and maintained by dedicated groups.

Some are explicitly promoted as hacking aids, while others are framed as dual-use tools for penetration testing, reflecting the broader “dual-use dilemma” in advanced technologies. “Any tool powerful enough to build a complex system can also be repurposed to break one,” the analysis by Palo Alto Networks’ Unit 42 noted.

Sophisticated language capabilities

WormGPT, created by blackhat developers, has returned in a more commercialised form, offering subscriptions reportedly priced at about $50 per month or $220 for lifetime access. Advertised across Telegram, forums, and its own interface, the latest version claims sophisticated language capabilities that can mimic executives or vendors, removing tell-tale grammatical errors common in traditional phishing.

In tests, researchers said WormGPT could generate functional PowerShell ransomware and convincing ransom notes featuring threats such as 72-hour payment deadlines and price doubling upon expiry.

KawaiiGPT 2.5, another model cited by researchers, has been open-sourced on GitHub since July and promotes a lightweight Linux setup that takes “less than five minutes.”

Tongue-in-cheek branding

It is backed by a community of roughly 500 contributors and presented with tongue-in-cheek branding as “Your Sadistic Cyber Pentesting Waifu,” underscoring how some projects straddle the line between offensive tooling and purported security research.

While these models can accelerate low-skill cyber operations, researchers cautioned that much of the code they produce remains relatively basic and is often detectable by modern security tools.

However, they warned that continued development and commercialisation could lower the barrier to entry for inexperienced attackers and increase the quality and volume of social engineering campaigns.

The trend parallels broader shifts in real-world attacks, where automation and AI have begun to handle substantial portions of intrusion workflows.

Analysts said defenders should expect more polished phishing at scale and quicker prototyping of commodity malware, and urged organisations to harden identity controls, email authentication, and script execution policies as these underground tools evolve.

Intel denies TSMC trade-secret allegations against returning official

  • TSMC alleged “a high probability that Lo uses, leaks, discloses or transfers TSMC’s trade secrets and confidential information to Intel,” saying legal action was necessary.
  • Taiwan’s economy ministry says will cooperate to determine whether the case involves infringement of core technologies or violations of the National Security Act.

Intel rejected allegations from Taiwan Semiconductor Manufacturing Co. (TSMC) that executive Wei-Jen Lo, a former senior vice president at the Taiwanese chipmaker, leaked trade secrets after joining Intel in October.

“Based on everything we know, we have no reason to believe there is any merit to the allegations involving Mr. Lo,” Intel said in an emailed statement, adding that it maintains “rigorous policies and controls” prohibiting the use or transfer of third-party confidential information or intellectual property.

Intel said it has welcomed Lo back and described him as widely respected for “integrity, leadership and technical expertise,” noting that talent movement across companies is “a common and healthy part of our industry.”

Effort to regain process leadership

TSMC said Tuesday it filed a lawsuit in Taiwan’s Intellectual Property and Commercial Court against Lo, who retired from TSMC after a 21-year tenure during which he helped drive mass production of cutting-edge 5-nanometre, 3-nm, and 2-nm chips. Lo previously worked at Intel for 18 years before joining TSMC in 2004.

In its statement, TSMC alleged “a high probability that Lo uses, leaks, discloses or transfers TSMC’s trade secrets and confidential information to Intel,” saying legal action was necessary. Taiwan’s economy ministry said it will cooperate to determine whether the case involves infringement of core technologies or violations of the National Security Act.

The legal dispute comes amid intensifying competition in advanced process nodes, with global chipmakers jockeying for leadership in 2-nm and below technologies. Intel’s recruitment of veteran foundry talent, including Lo, is central to its effort to regain process leadership and scale its contract manufacturing business, while TSMC seeks to protect its IP and market dominance.