Competition regulation

Chen Deli warns that AGI progress could bring dangerous societal consequences

DeepSeek made a rare public appearance at the World Internet Conference in Wuzhen, where senior researcher Chen Deli restated the firm’s ambition to develop AGI. He joined other companies known as China’s ‘six little dragons’ of AI and acknowledged the potential risks of advanced systems.

Chen represented founder Liang Wenfeng, who has remained out of the public eye since meeting President Xi Jinping earlier this year. He said AI’s current limits create a short ‘honeymoon phase’ before automation reshapes employment and social stability.

The start-up, founded in 2023 as a High-Flyer spin-out, continues to focus on long-term AGI research rather than short-lived commercial trends. Chen said it was reasonable to consider the dangers of highly capable systems while still pursuing them.

His comments echoed an open letter calling for a pause on superintelligence work until strong public support and scientific consensus on safety emerge. Hundreds of experts and public figures backed the appeal for tighter oversight.

Chen argued that market incentives make slowing progress unrealistic and said widespread job replacement may ultimately define the AI revolution. Other firms from China, including Zhipu AI and Alibaba, outlined plans for more powerful infrastructure to meet rising compute demand.

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

UK unveils major push to drive national AI growth

A significant wave of public and private investment is set to place AI at the centre of the UK’s growth strategy. AI Growth Zones backed by substantial investment will drive job creation, high-tech infrastructure and local industry development across regions such as South Wales, London and Bristol.

Government officials stated that the measures aim to provide British firms with the tools necessary to scale and compete globally.

South Wales will host a significant £10 billion development expected to create over 5,000 jobs in the next decade. The zone will focus on data centres, advanced computing and AI research, supported by government funding for skills development and business adoption.

International tech companies expanding in the UK include Microsoft, Vantage Data Centres, Groq and Perplexity AI, each committing to new sites and enlarged workforces.

Further support will expand access to computing for researchers and start-ups nationwide. A government-backed advance market commitment worth up to £100 million will help hardware-focused AI firms secure their first key customers.

Officials confirmed nearly £500 million for the Sovereign AI Unit, which will scale domestic capabilities and back high-potential firms. Up to £137 million will also support the UK’s new AI-for-science strategy, which focuses on accelerating drug discovery and other breakthroughs.

Government representatives and industry leaders described the announcements as a turning point for the UK’s innovation capacity. Supporters say the measures will strengthen Britain’s tech leadership while creating jobs, boosting regional economies and advancing scientific progress.

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

Tech groups welcome EU reforms as privacy advocates warn of retreat

The EU has unveiled plans to scale back certain aspects of its AI and data privacy rules to revive innovation and alleviate regulatory pressure on businesses. The Digital Omnibus package delays stricter oversight for high-risk AI until 2027 and permits the use of anonymised personal data for model training.

The reforms amend the AI Act and several digital laws, cutting cookie pop-ups and simplifying documentation requirements for smaller firms. EU tech chief Henna Virkkunen says the aim is to boost competitiveness by removing layers of rigid regulation that have hindered start-ups and SMEs.

US tech lobby groups welcomed the overall direction. Still, they criticised the package for not going far enough, particularly on compute thresholds for systemic-risk AI and copyright provisions with cross-border effects. They argue the reforms only partially address industry concerns.

Privacy and digital rights advocates sharply opposed the changes, warning they represent a significant retreat from Europe’s rights-centric regulatory model. Groups including NOYB accused Brussels of undermining hard-won protections in favour of Big Tech interests.

Legal scholars say the proposals could shift Europe closer to a more permissive, industry-driven approach to AI and data use. They warn that the reforms may dilute the EU’s global reputation as a standard-setter for digital rights, just as the world seeks alternatives to US-style regulation.

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

EU eases AI and data rules to boost tech growth

The European Commission has proposed easing AI and data privacy rules to cut red tape and help European tech firms compete internationally. Companies could access datasets more freely for AI training and have 16 months to comply with ‘high-risk’ AI rules.

Brussels also aims to cut the number of cookie pop-ups, allowing users to manage consent more efficiently while protecting privacy. The move has sparked concern among rights groups and campaigners who fear the EU may be softening its stance on Big Tech.

Critics argue that loosening regulations could undermine citizen protections, while European companies welcome the changes as a way to foster innovation and reduce regulatory burdens that have slowed start-ups and smaller businesses.

EU officials emphasise that the reforms seek a balance between competitiveness and safeguarding fundamental rights. Commission officials say the measures will help European firms compete with US and Chinese rivals while safeguarding citizen privacy.

Simplifying consent mechanisms and providing companies more operational flexibility are central to the plan’s goals.

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

US administration pushes back on proposal to restrict Nvidia sales to China

The White House is urging Congress to reject a bipartisan proposal that would restrict Nvidia from selling advanced AI chips to China and other countries subject to an embargo. The GAIN AI Act would require chipmakers to prioritise US buyers before exporting high-performance hardware.

Lawmakers are debating whether to attach the provision to the annual defence spending bill, a move that could accelerate approval. The White House intervention represents a significant win for Nvidia, which has lobbied to maintain export flexibility amid shifting trade policies.

China was previously a significant market for Nvidia, but the firm has pared back expectations due to rising geopolitical risks. Beijing has also increased scrutiny of US-made chips as it pushes for self-reliance in AI and semiconductor technology.

The policy discussions come shortly after Nvidia posted stronger-than-expected third-quarter earnings and issued an upbeat outlook. CEO Jensen Huang has pushed back against concerns of an AI-driven valuation bubble, arguing demand remains robust.

Nvidia’s shares rose 5 percent after hours following the earnings report, reflecting investor confidence as Washington continues to debate the future of AI chip export controls.

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

AI in healthcare gains regulatory compass from UK experts

Professor Alastair Denniston has outlined the core principles for regulating AI in healthcare, describing AI as the ‘X-ray moment’ of our time.

Like previous innovations such as MRI scanners and antibiotics, AI has the potential to improve diagnosis, treatment and personalised care dramatically. Still, it also requires careful oversight to ensure patient safety.

The MHRA’s National Commission on the Regulation of AI in Healthcare is developing a framework based on three key principles. The framework must be safe, ensuring proportionate regulation that protects patients without stifling innovation.

It must be fast, reducing delays in bringing beneficial technologies to patients and supporting small innovators who cannot endure long regulatory timelines. Ultimately, it must be trusted, with transparent processes that foster confidence in AI technologies today and in the future.

Professor Denniston emphasises that AI is not a single technology but a rapidly evolving ecosystem. The regulatory system must keep pace with advances while allowing the NHS to harness AI safely and efficiently.

Just as with earlier medical breakthroughs, failure to innovate can carry risks equal to the dangers of new technologies themselves.

The National Commission will soon invite the public to contribute their views through a call for evidence.

Patients, healthcare professionals, and members of the public are encouraged to share what matters to them, helping to shape a framework that balances safety, speed, and trust while unlocking the full potential of AI in the NHS.

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

Target expands OpenAI partnership with new ChatGPT shopping app

Target is expanding its partnership with OpenAI by launching a new shopping app directly inside ChatGPT. The app offers customers personalised recommendations, multi-item baskets and streamlined checkout across Drive Up, Order Pickup and shipping.

The retailer will continue using OpenAI’s models and ChatGPT Enterprise to enhance employee productivity and strengthen digital experiences across its business.

AI is central to Target’s operations, supporting supply-chain forecasts, store processes, and personalised digital tools. Over 18,000 employees utilise ChatGPT Enterprise to streamline routine tasks, enhance creativity, and receive faster support for guest requests and returns through internal AI assistants.

Customer-facing tools such as Shopping Assistant, Gift Finder, Guest Assist and JOY reinforce this strategy by offering curated suggestions and instant answers.

The new Target app inside ChatGPT extends this AI-driven approach to customers. Shoppers will be able to ask for ideas, browse curated suggestions, build baskets and check out through their Target accounts.

The beta version launches next week, and upcoming features include Target Circle linking and same-day delivery. Target views the partnership as part of a retail shift, embedding AI across products, operations and guest interactions to drive the next wave of innovation.

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

EU introduces plan to strengthen consumer protection

The European Commission has unveiled the 2030 Consumer Agenda, a strategic plan to reinforce protection, trust, and competitiveness across the EU.

With 450 million consumers contributing over half of the Union’s GDP, the agenda aims to simplify administrative processes for businesses, rather than adding new burdens, while ensuring fair treatment for shoppers.

The agenda sets four priorities to adapt to rising living costs, evolving online markets, and the surge in e-commerce. Completing the Single Market will remove cross-border barriers, enhance travel and financial services, and evaluate the effectiveness of the Geo-Blocking Regulation.

A planned Digital Fairness Act will address harmful online practices, focusing on protecting children and strengthening consumer rights.

Sustainable consumption takes a central focus, with efforts to combat greenwashing, expand access to sustainable goods, and support circular initiatives such as second-hand markets and repairable products.

The Commission will also enhance enforcement to tackle unsafe or non-compliant products, particularly from third countries, ensuring that compliant businesses are shielded from unfair competition.

Implementation will be overseen through the Annual Consumer Summit and regular Ministerial Forums, which will provide political guidance and monitor progress.

The 2030 Consumer Agenda builds on prior achievements and EU consultations, aiming to modernise consumer protection instead of leaving gaps in a rapidly changing market.

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

Sundar Pichai warns users not to trust AI tools easily

Google CEO Sundar Pichai advises people not to unquestioningly trust AI tools, warning that current models remain prone to errors. He told the BBC that users should rely on a broader information ecosystem rather than treat AI as a single source of truth.

Pichai said generative systems can produce inaccuracies and stressed that people must learn what the tools are good at. The remarks follow criticism of Google’s own AI Overviews feature, which attracted attention for erratic and misleading responses during its rollout.

Experts say the risk grows when users depend on chatbots for health, science, or news. BBC research found major AI assistants misrepresented news stories in nearly half of the tests this year, underscoring concerns about factual reliability and the limits of current models.

Google is launching Gemini 3.0, which it claims offers stronger multimodal understanding and reasoning. The company says its new AI Mode in search marks a shift in how users interact with online information, as it seeks to defend market share against ChatGPT and other rivals.

Pichai says Google is increasing its investment in AI security and releasing tools to detect AI-generated images. He maintains that no single company should control such powerful technology and argues that the industry remains far from a scenario in which one firm dominates AI development.

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

Europe’s digital sovereignty advances through SAP’s new AI collaborations

SAP has announced new partnerships with Bleu, Capgemini, and Mistral AI to advance Europe’s digital sovereignty. The collaboration combines SAP’s expertise in enterprise software with France’s AI ecosystem to develop secure, scalable, and sovereign cloud solutions for governments and regulated sectors.

Bleu and Delos Cloud have established a Franco-German alliance focused on crisis resilience, creating joint capabilities for early detection, analysis, and remediation of cyber incidents. Their cooperation supports rapid response in extreme scenarios and reinforces control over critical infrastructure.

SAP and Capgemini are expanding their partnership to advance sovereign agentic AI and strengthen cybersecurity across Europe. Their new Sovereign Technology Partnership will deliver data management, cloud services, and automation tools for public and regulated sectors.

SAP and Mistral AI are also deepening their collaboration to create Europe’s first full sovereign AI stack. SAP will offer Mistral’s frontier models through its sovereign AI foundation on SAP BTP, while both companies co-develop industry-specific AI applications designed for engineering and R&D workloads.

These partnerships form part of SAP’s broader sovereign cloud strategy, backed by more than €20bn in investment. SAP states that its aim is to provide a secure, compliant, and locally controlled infrastructure that enables innovation while safeguarding European data, assets, and long-term technological independence.

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