Digital business models

Google DeepMind updates AI safety framework for advanced risks

A leading AI developer has released the third iteration of its Frontier Safety Framework (FSF), aiming to identify and mitigate severe risks from advanced AI models. The update expands risk domains and refines the process for assessing potential threats.

Key changes include the introduction of a Critical Capability Level (CCL) focused on harmful manipulation. The update targets AI models with the potential to systematically influence beliefs and behaviours in high-stakes contexts, ensuring safety measures keep pace with growing model capabilities.

The framework also enhances protocols for misalignment risks, addressing scenarios where AI could override operators’ control or shutdown attempts. Safety case reviews are now conducted before external launches and large-scale internal deployments reach critical thresholds.

The updated FSF sharpens risk assessments and applies safety and security mitigations in proportion to threat severity. It reflects a commitment to evidence-based AI governance, expert collaboration, and ensuring AI benefits humanity while minimising risks.

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

UAE commits to global crypto tax transparency

The United Arab Emirates has signed the Multilateral Competent Authority Agreement on the Automatic Exchange of Information under the Crypto-Asset Reporting Framework (CARF). The move follows the Ministry of Finance’s earlier commitment to strengthen oversight of digital assets.

Implementation of CARF in the UAE is scheduled for 2027, with the first exchange of information expected in 2028. The framework will enable automatic sharing of crypto tax data, providing greater certainty for the sector and reinforcing global transparency.

Authorities are calling on stakeholders across the crypto-asset ecosystem, including intermediaries, exchanges, custodians and traders, to contribute to a public consultation on the implementation of CARF.

The consultation, which opened on 15 September 2025 and runs until 8 November 2025, seeks expert input to help shape regulatory rules aligned with market needs.

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

Lawmakers told to harness AI for democracy

The Speaker of the ECOWAS Parliament, Hajia Memounatou Ibrahima, has urged lawmakers to embrace AI to strengthen legislative work and democratic governance across West Africa.

Speaking at the Parliament’s 2025 Second Extraordinary Session in Port Harcourt, she said AI could improve communication with citizens, support regulatory frameworks, and drive inclusive growth.

Nigeria’s Deputy Senate President, Sen. Jibrin Barau, highlighted AI’s potential to enhance efficiency, accountability and oversight while cautioning lawmakers to examine associated risks.

Nigerian ambassador Ahmed Dunoma, Permanent Secretary of Ministry of Foreign Affairs, stressed that AI is already shaping education and security sectors. He called for proactive guidance to ensure its deployment deepens integration and safeguards democracy.

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

Amazon outlines responsible AI and global internet plans at UN

Amazon is meeting world leaders at the 80th UN General Assembly to share its vision for responsible AI and global internet access. The company highlighted Project Kuiper’s satellite initiative to provide affordable internet to underserved communities and bridge the digital divide.

The initiative aims to deliver fast, affordable internet to communities without access, boosting education and economic opportunities. Connectivity is presented as essential for participation in the modern economy, as well as for cultural and knowledge exchange across the globe.

Amazon emphasised the development of AI tools that are responsible, inclusive, and designed to enhance human potential. The company aims to make technology accessible, helping small businesses, speeding research, and offering tools once reserved for large organisations.

Collaboration remains central to Amazon’s approach. The company plans to work with governments, the UN, civil society, and other private sector partners to ensure technological advancements benefit humanity while mitigating potential risks.

Discussions at UNGA80 are expected to shape future strategies for innovation, governance, and sustainable development.

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

Jaguar shutdown extended as ministers meet suppliers

Jaguar Land Rover (JLR) has confirmed its factories will remain closed until at least 1 October, extending a shutdown triggered by a cyber-attack in late August.

Business Secretary Peter Kyle and Industry Minister Chris McDonald are meeting JLR and its suppliers, as fears mount that small firms in the supply chain could collapse without the support of the August cyberattack.

The disruption, estimated to cost JLR £50m per week, affects UK plants in Solihull, Halewood and Wolverhampton. About 30,000 people work directly for JLR, with a further 100,000 in its supply chain.

Unions say some supplier staff have been laid off with little or no pay, forcing them to seek Universal Credit. Unite has called for a furlough-style scheme, while MPs have pressed the government to consider emergency loans.

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

Europe prepares formal call for AI Gigafactory projects

The European Commission is collaborating with the EU capitals to narrow the list of proposals for large AI training hubs, known as AI Gigafactories. The €20 billion plan will be funded by the Commission (17%), the EU countries (17%), and industry (66%) to boost computing capacity for European developers.

The first call drew 76 proposals from 16 countries, far exceeding the initially planned four or five facilities. Most submissions must be merged or dropped, with Poland already seeking a joint bid with the Baltic states as talks continue.

Some EU members will inevitably lose out, with Ursula von der Leyen, the President of the European Commission, hinting that priority could be given to countries already hosting AI Factories. That could benefit Finland, whose Lumi supercomputer is part of a Nokia-led bid to scale up into a Gigafactory.

The plan has raised concerns that Europe’s efforts come too late, as US tech giants invest heavily in larger AI hubs. Still, Brussels hopes its initiative will allow EU developers to compete globally while maintaining control over critical AI infrastructure.

A formal call for proposals is expected by the end of the year, once the legal framework is finalised. Selection criteria and funding conditions will be set to launch construction as early as 2026.

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

BlackRock backs South Korea push to become Asia AI hub

South Korea has secured a significant partnership with BlackRock to accelerate its ambition of becoming Asia’s leading AI hub. The agreement will see the global asset manager join the Ministry of Science and ICT in developing hyperscale AI data centres.

A deal that followed a meeting between President Lee Jae Myung and BlackRock chair Larry Fink, who pledged to attract large-scale international investment into the country’s AI infrastructure.

Although no figures were disclosed, the partnership is expected to focus on meeting rising demand from domestic users and the wider Asia-Pacific region, with renewable energy powering the facilities.

The move comes as Seoul increases national funding for AI, semiconductors and other strategic technologies to KRW150 trillion ($107.7 billion). South Korean companies are also stepping up efforts, with SK Telecom announcing plans to raise AI investment to a third of its revenue over five years.

BlackRock’s involvement signals international confidence in South Korea’s long-term vision to position itself as a regional AI powerhouse and secure a leadership role in next-generation digital infrastructure.

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

Xueba 01 becomes first robot doctoral student in China

Standing 1.75 metres tall and weighing 32 kilograms, Xueba 01 has become the first robot doctoral student at the Shanghai Theatre Academy. Over the next four years, it will study digital performance design, focusing on traditional Chinese opera movements and techniques.

The programme, launched in partnership with the University of Shanghai for Science and Technology, combines technical and artistic training. USST provides technical guidance, while STA develops the robot’s artistic performance, focusing on interaction, expression, and cognitive growth.

Xueba 01 features an advanced tendon-based bionic structure, human-like facial technology, and the ability to perform over 100 lifelike expressions. Based on audience feedback, it can adjust its height and appearance, perform for extended periods, and adapt its performance in real time.

Motion capture technology helps it learn from professional performers to refine movements and gestures.

STA faculty highlight the robot’s role in exploring the intersection of art and technology. The initiative aims to integrate AI with traditional Chinese arts, preserve cultural heritage, and inspire contemporary artists to combine technological literacy with humanistic and interdisciplinary skills.

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

Oracle to oversee TikTok algorithm in US deal

The White House has confirmed that TikTok’s prized algorithm will be managed in the US under Oracle’s supervision as part of a deal to place the app’s US operations under majority American ownership. The agreement would transfer control of TikTok’s US business, along with a copy of the algorithm, to a new joint venture run by a board dominated by American investors.

The confirmed participants are Oracle and private equity firm Silver Lake, with Fox Corp. also expected to join the group. President Donald Trump has suggested that high-profile figures such as Michael Dell, Rupert, and Lachlan Murdoch could be involved, though CNN sources say that the Murdochs personally will not invest. ByteDance will keep a stake of less than 20% in the new US entity.

The deal follows years of negotiations over concerns that TikTok’s Chinese parent company could be pressured to manipulate the platform for political influence. By law, ByteDance is barred from cooperating on the algorithm with any new American owners. The code will be reviewed, retrained on US user data to address these fears, and monitored by Oracle to ensure its independence.

President Trump is expected to sign an executive order later this week certifying that the deal meets national security requirements under last year’s ‘ban-or-sale’ law. He will also extend the pause on enforcement by 120 days, giving Washington and Beijing time to finalise regulatory approvals. The White House said the deal could be signed within days, with completion likely early next year.

The arrangement deepens Oracle’s role in managing TikTok’s American presence, building on its existing partnership to store US user data. The development coincided with Oracle announcing a leadership shake-up, with CEO Safra Catz stepping down to become vice chair and two co-CEOs taking over. It is unclear if the timing is connected, but Catz, a close Trump ally, could take a role in the TikTok venture.

While financial details remain uncertain, the White House has ruled out taking a direct stake in the company. The deal, valued in the billions, would conclude a years-long effort to bring TikTok under US oversight and resolve national security concerns tied to its Chinese ownership.

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

Misconfigurations drive major global data breaches

Misconfigurations in cloud systems and enterprise networks remain one of the most persistent and damaging causes of data breaches worldwide.

Recent incidents have highlighted the scale of the issue, including a cloud breach at the US Department of Homeland Security, where sensitive intelligence data was inadvertently exposed to thousands of unauthorised users.

Experts say such lapses are often more about people and processes than technology. Complex workflows, rapid deployment cycles and poor oversight allow errors to spread across entire systems. Misconfigured servers, storage buckets or access permissions then become easy entry points for attackers.

Analysts argue that preventing these mistakes requires better governance, training and process discipline rather. Building strong safeguards and ensuring staff have the knowledge to configure systems securely are critical to closing one of the most exploited doors in cybersecurity.

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