Artificial intelligence

Meta faces fresh EU backlash over Digital Markets Act non-compliance

Meta is again under EU scrutiny after failing to fully comply with the bloc’s Digital Markets Act (DMA), despite a €200 million fine earlier this year.

The European Commission says Meta’s current ‘pay or consent’ model still falls short and could trigger further penalties. A formal warning is expected, with recurring fines likely if the company does not adjust its approach.

The DMA imposes strict rules on major tech platforms to reduce market dominance and protect digital fairness. While Meta claims its model meets legal standards, the Commission says progress has been minimal.

Over the past year, Meta has faced nearly €1 billion in EU fines, including €798 million for linking Facebook Marketplace to its central platform. The new case adds to years of tension over data practices and user consent.

The ‘pay or consent’ model offers users a choice between paying for privacy or accepting targeted ads. Regulators argue this does not meet the threshold for genuine consent and mirrors Meta’s past GDPR tactics.

Privacy advocates have long criticised Meta’s approach, saying users are left with no meaningful alternatives. Internal documents show Meta lobbied against privacy reforms and warned governments about reduced investment.

The Commission now holds greater power under the DMA than it did with GDPR, allowing for faster, centralised enforcement and fines of up to 10% of global turnover.

Apple has already been fined €500 million, and Google is also under investigation. The EU’s rapid action signals a stricter stance on platform accountability. The message for Meta and other tech giants is clear: partial compliance is no longer enough to avoid serious regulatory consequences.

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

Broadcom challenges Nvidia with Tomahawk Ultra AI networking chip

Broadcom has introduced a new networking chip designed to boost AI data centre performance, positioning itself against Nvidia in the AI hardware race.

Called Tomahawk Ultra, the chip helps link hundreds of processors within close range, ensuring fast communication vital for large AI models.

Instead of using Nvidia’s proprietary NVLink system, Broadcom’s Tomahawk Ultra operates on an accelerated version of Ethernet and connects up to four times more chips within a server rack.

The chip acts as a traffic controller, enabling what the industry refers to as ‘scale-up’ computing — where AI models tap into combined computing power from tightly grouped chips.

According to Broadcom senior vice president Ram Velaga, it took engineers around three years to design the Tomahawk Ultra, which was originally aimed at high-performance computing but adapted for AI workloads as demand surged.

Taiwan Semiconductor Manufacturing Company (TSMC) is producing the processors using its five-nanometre process, and the chips are already shipping to customers.

Rather than requiring major hardware overhauls, the Tomahawk Ultra offers a scalable, Ethernet-based alternative for AI firms looking to build faster, more flexible data centres.

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

UK customer‑experience leaders champion human‑centred AI

Marketing Week’s CX50 2025 accolades spotlight ten individuals who are redefining customer experience by putting people before pixels.

These leaders are harnessing AI thoughtfully, augmenting frontline staff with tools that personalise service without eroding the human touch.

Their work spans retail, hospitality, financial services and the public sector, showcasing how AI-driven systems can streamline operations while nurturing empathetic engagement.

Strategies include using automation for routine tasks and intelligent routing, freeing teams to focus on empathy, insight and dialogue.

Partners Cognizant, Google Cloud and Salesforce helped assemble the CX50 list. The initiative underscores a growing movement: top-performing brands blend algorithmic precision with human warmth to build trust and loyalty in increasingly digital markets.

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

AI system screens diabetic eye disease with near-perfect accuracy

A new AI programme is showing remarkable accuracy in detecting diabetic retinopathy, a leading cause of preventable blindness. The SMART system, short for Simple Mobile AI Retina Tracker, can scan retinal images using even basic smartphones and has achieved over 99% accuracy.

Researchers at the University of Texas Health Sciences Center in the US trained the AI using thousands of retinal images from diverse populations across six continents. The system processes images in under a second and can distinguish diabetic retinopathy from other eye diseases.

Experts say the technology could dramatically expand access to eye screenings, particularly in areas lacking specialist care. By integrating the tool into regular check-ups, both primary care providers and ophthalmologists could streamline early diagnosis.

Researchers highlighted that the tool’s mobile accessibility allows for global reach, potentially screening billions. The findings were presented at the annual meeting of The Endocrine Society, though they have yet to be peer-reviewed.

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

Tech giants work to avert an AI‑driven energy crisis

The AI boom is triggering alarms in the energy sector, with data centres expected to consume 3% of the world’s electricity by 2030, double today’s share. This projection has spurred tech firms to pursue aggressive efficiency strategies.

Control systems powered by AI regulate cooling more precisely, while liquid cooling systems replace inefficient air-based approaches. Cutting-edge chips and more innovative software have already yielded energy savings of 20–30% per workflow.

Infrastructure improvements mean data-centre support systems now consume just 10% of the energy that powering the servers does. Despite these advances, the total energy demand continues to rise, prompting investments in technology upgrades and low-carbon power sources.

A global race is underway as US and Chinese companies push for increasingly efficient AI chips. Still, experts warn that overall consumption will grow, so efficiency alone won’t be enough. Broad energy planning and sustainable strategies are critical to avoid a looming power crunch as AI proliferates.

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

Pennsylvania criminalises malicious deepfakes under new digital forgery law

Governor Shapiro has enacted a new statute enhancing Pennsylvania’s legal stance on AI-generated content by defining deceptive deepfakes as digital forgery.

The law criminalises creating and distributing such content, mainly when used for deceit, highlighting a proactive response to deepening online threats.

The legislation differentiates between uses of deepfakes: non-consensual impersonation will result in misdemeanour charges, while cases involving fraudulent intent, such as financial scams or political manipulation, are now classified as third-degree felonies.

Support for the bill was bipartisan and overwhelming in the state legislature. Its sponsors emphasised that while it deters harmful digital impersonation, it also carefully safeguards legitimate speech, including parody, satire, and artistic expression.

With Pennsylvania now among the growing number of states implementing deepfake regulations, this development aligns with a national trend to regulate AI-generated digital forgeries. It complements earlier state-level laws and federal initiatives to curb AI’s misuse without stifling innovation.

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

AI tool uses walking patterns to detect early signs of dementia

Fujitsu and Acer Medical are trialling an AI-powered tool to help identify early signs of dementia and Parkinson’s disease by analysing patients’ walking patterns. The system, called aiGait and powered by Fujitsu’s Uvance skeleton recognition technology, converts routine movements into health data.

Initial tests are taking place at a daycare centre linked to Taipei Veterans Hospital, using tablets and smartphones to record basic patient movements. The AI compares this footage with known movement patterns associated with neurodegenerative conditions, helping caregivers detect subtle abnormalities.

The tool is designed to support early intervention, with abnormal results prompting follow-up by healthcare professionals. Acer Medical plans to expand the service to elderly care centres across Taiwan by the end of the year.

Fujitsu’s AI was originally developed for gymnastics scoring and adapted to analyse real-world gait data with high accuracy using everyday mobile devices. Both companies hope to extend the technology’s use to paediatrics, sports science, and rehabilitation in future.

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

South Korean firm unveils faster AI data centre architecture with CXL-over-Xlink

South Korean company Panmnesia has introduced a new architecture for AI data centres aimed at improving speed and efficiency.

Instead of using only PCIe or RDMA-based systems, its CXL-over-Xlink approach combines Compute Express Link (CXL) with fast accelerator links such as UALink and NVLink.

The company claims this design can deliver up to 5.3 times faster AI training and reduce inference latency sixfold. By allowing CPUs and GPUs to access large shared memory pools via the CXL fabric, AI workloads are no longer restricted by the fixed memory limits inside each GPU.

It will enable data centres to scale compute and memory independently, adapting to changing workload demands without hardware overprovisioning.

Panmnesia’s system also reduces communication overhead using accelerator-optimised links for CXL traffic, helping maintain high throughput with sub-100ns latency.

The architecture incorporates a hierarchical memory model blending local high-bandwidth memory with pooled CXL memory, alongside scalable CXL 3.1 switches that connect hundreds of devices efficiently without bottlenecks.

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

Thinking Machines Lab raises $2bn to build safer AI

Thinking Machines Lab, an AI startup founded earlier this year by former OpenAI executive Mira Murati, has raised $2 billion in new funding. The round, which values the company at $12 billion, was led by Andreessen Horowitz and backed by Nvidia, Cisco, AMD, and others.

The company aims to develop safer and more reliable AI systems by focusing on how people naturally interact with the world, including speech and vision. Its first product, due in the coming months, will offer open-source components designed to support researchers and startups.

At launch, nearly two-thirds of the team had previously worked at OpenAI, underscoring the company’s ambition to lead in the field of frontier AI. Murati said the startup plans to make its science publicly available to support understanding and transparency.

The investment comes amid a surge in AI-related funding, which accounted for over 64% of all US startup deal value in the first half of 2025. Growing interest in generative and multimodal AI continues to attract major capital despite wider concerns over tech sector spending.

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

AI Appreciation Day highlights progress and growing concerns

AI is marking another milestone as experts worldwide reflect on its rapid rise during AI Appreciation Day. From reshaping business workflows to transforming customer experiences, AI’s presence is expanding — but so are concerns over its long-term implications.

Industry leaders point to AI’s growing role across sectors. Patrick Harrington from MetaRouter highlights how control over first-party data is now seen as key instead of just processing large datasets.

Vall Herard of Saifr adds that successful AI implementations depend on combining curated data with human oversight rather than relying purely on machine-driven systems.

Meanwhile, Paula Felstead from HBX Group believes AI could significantly enhance travel experiences, though scaling it across entire organisations remains a challenge.

Voice AI is changing industries that depend on customer interaction, according to Natalie Rutgers from Deepgram. Instead of complex interfaces, voice technology is improving communication in restaurants, hospitals, and banks.

At the same time, experts like Ivan Novikov from Wallarm stress the importance of securing AI systems and the APIs connecting them, as these form the backbone of modern AI services.

While some celebrate AI’s advances, others raise caution. SentinelOne’s Ezzeldin Hussein envisions AI becoming a trusted partner through responsible development rather than unchecked growth.

Naomi Buckwalter from Contrast Security warns that AI-generated code could open security gaps instead of fully replacing human engineering, while Geoff Burke from Object First notes that AI-powered cyberattacks are becoming inevitable for businesses unable to keep pace with evolving threats.

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