Digital Watch newsletter – Issue 97 – March 2025

 Hourglass

Snapshot: The developments that made waves

AI governance

Chinese companies are increasingly backing DeepSeek‘s AI, marking a pivotal moment for the industry.

The European Commission has launched the OpenEuroLLM Project, a new initiative to develop open-source, multilingual AI models.

Australia has banned Chinese AI startup DeepSeek from all government devices, citing security risks.

World leaders gathered in Paris for the second day of the Artificial Intelligence (AI) Action Summit, where the focus turned to balancing national interests with global cooperation.

In his op-ed, From Hammurabi to ChatGPT, Jovan Kurbalija draws on the ancient Code of Hammurabi to argue for a principle of legal accountability in modern AI regulation and governance.

Technologies

DeepSeek, an up-and-coming AI startup from China, is rapidly pushing forward the release of its latest AI model, R2, following the remarkable success of its predecessor, R1.

Elon Musk’s AI startup, xAI, has unveiled its latest AI model, Grok-3, which the billionaire claims is the most advanced chatbot technology.

The New York Times has officially approved the use of AI tools for its editorial and product teams, allowing AI to assist with tasks such as generating social media copy, writing SEO headlines, and coding.

Microsoft has announced a groundbreaking quantum computing chip, Majorana 1, which it claims could make useful quantum computers a reality within years.

China has warned that the United States‘ efforts to pressure other nations into targeting its semiconductor industry will ultimately backfire.

Infrastructure

A state-of-the-art space lab on the outskirts of Cairo, touted as Africa’s first satellite production facility, has been built with substantial Chinese involvement. 

Elon Musk’s Starlink network is facing increasing competition in the satellite internet market, particularly from SpaceSail, a Shanghai-based company backed by the Chinese government, and Amazon’s Project Kuiper.

Cybersecurity

The EU Commission introduced a proposal aimed at strengthening the EU’s response to large-scale cyber attacks.

Hackers have stolen $1.5 billion from Dubai-based cryptocurrency exchange Bybit in what is believed to be the largest digital heist in history.

Following the recent security breach at Bybit, major cryptocurrency firms have joined forces to combat the attack and mitigate its impact.

Digital rights

Elon Musk has reignited his rivalry with OpenAI by leading a consortium in a staggering $97.4 billion bid to acquire the nonprofit that governs the ChatGPT creator.

South Korea’s National Intelligence Service (NIS) has raised concerns about the Chinese AI app DeepSeek, accusing it of excessively collecting personal data and using it for training purposes.

TikTok has introduced a new method for US Android users to download the app directly from its website, bypassing restrictions imposed by app stores.

South Korea’s data protection authority has suspended new downloads of the Chinese AI app DeepSeek, citing concerns over non-compliance with the country’s privacy laws.

A US federal judge has denied a request to temporarily block Elon Musk’s Department of Government Efficiency (DOGE) from accessing data from seven federal agencies or making further workforce cuts.

Legal

An online education company has filed a lawsuit against Google, claiming its AI-generated search overviews are damaging digital publishing.

The Trump administration is reevaluating the conditions of CHIPS and Science Act subsidies, which allocate $39 billion to boost domestic semiconductor production.

Elon Musk-owned social media platform X has successfully removed a judge from a German court case concerning demands for real-time election data.

Trump Media & Technology Group and Rumble have filed an emergency motion in a US court against Brazilian Supreme Court Justice Alexandre de Moraes.

Internet economy

Chinese investors are flocking to AI-related stocks, betting that the success of home-grown startup DeepSeek will propel China to the forefront of the AI race amid the escalating Sino-US technology conflict. 

In a meeting at the White House on Friday, US President Donald Trump and Nvidia CEO Jensen Huang discussed the emerging challenges posed by China’s AI advancements, particularly the rapid rise of DeepSeek that has disrupted the global tech industry.

Macron announced private sector investments totalling around 109 billion euros ($112.5 billion) in its AI sector at the Paris AI summit, according to President Emmanuel Macron.

Elon Musk, who is leading Donald Trump’s federal cost-cutting initiative, has announced that efforts are underway to shut down the United States Agency for International Development (USAID).

President Donald Trump confirmed on Wednesday that he was in active discussions with China over the future of TikTok, as the US seeks to broker a sale of the popular app.

Elon Musk’s social media company X is currently discussing raising funds from investors at a $44 billion valuation.

Development

Alibaba has announced plans to invest at least 380 billion yuan ($52.44 billion) in cloud computing and AI infrastructure over the next three years.

US drugmaker Amgen has announced a $200 million investment in a new technology centre in southern India, which will focus on using AI and data science to support the development of new medicines.

Sociocultural

A German court has ruled that Elon Musk’s social media platform X must provide researchers with data to track the spread of misinformation ahead of the country’s national election on 23 February.

French prosecutors have launched an investigation into X, formerly known as Twitter, over alleged algorithmic bias.

Meta has launched a new initiative with UNESCO to enhance AI language recognition and translation, focusing on underserved languages.

A recent report by Australia’s eSafety regulator reveals that children in the country are finding it easy to bypass age restrictions on social media platforms.

A Russian court has fined Google 3.8 million roubles (£32,600) for hosting YouTube videos that allegedly instructed Russian soldiers on how to surrender.

Mexico has strongly opposed Google’s decision to rename the Gulf of Mexico as the ‘Gulf of America’ for US Google Maps users.

For more information on cybersecurity, digital policies, AI governance and other related topics, visit diplomacy.edu.


Data Protection Day 2025: A new mandate for data protection

Data Protection Day 2025 brought experts to Brussels and online to discuss the latest privacy challenges. The event featured key speeches, panels, and discussions on data protection enforcement, international cooperation, and the impact of emerging technologies.

In his keynote, Leonardo Cervera Navas warned about algorithms being used to destabilise the EU democracies and the power imbalance caused by tech monopolies. Despite these challenges, he emphasised that European values and regulatory tools like the Digital Clearing House 2.0 could help uphold privacy.

AD 4nXdmpGPG AL98foAOOO5roALPZOPmpQMnQS9nGmVYA8TE1X3h V4FREpOLCvfOzlOhX84Av5lyytjnr1 03pJCSMPxJm8gepCjIT55 ZzeUMmfKcriOJt

A panel moderated by Kait Bolongaro examined data protection priorities and enforcement challenges. European Data Protection Supervisor Wojciech Wiewiorowski stressed the role of strong data protection authorities (DPAs) in ensuring GDPR compliance. Matthias Kloth highlighted the modernisation of Convention 108+ as a key legal advancement. The discussion also covered international cooperation, with panellists agreeing that while global collaboration is necessary, privacy standards must not be diluted.

Audience questions raised concerns about the complexity and cost of enforcing privacy rights. Wiewiorowski saw no major GDPR overhaul in the near future but suggested procedural adjustments. Jo Pierson proposed that civil society organisations assist individuals facing legal barriers. The discussion also touched on China’s DeepSeek AI, with panellists refraining from labelling it a threat but stressing the need for Europe to control AI training datasets. Wiewiorowski likened the AI race to the Cold War’s ‘Sputnik moment,’ questioning if Europe was falling behind.

A session on future data protection challenges, led by Anna Buchta, highlighted regulatory complexity. Marina Kaljurand noted a significant digital knowledge gap among European lawmakers, complicating policy decisions. She also raised concerns about the UK’s adequacy decision, which expires in June 2025, and worsening US-EU relations under Trump. UK Information Commissioner John Edwards emphasised AI’s disruptive impact on biometrics and tracking, while Alberto Di Felice from DIGITALEUROPE criticised excessive bureaucracy, advocating for streamlined regulatory oversight.

Beatriz de Anchorena, head of Argentina’s data protection authority, championed Convention 108+ as a global privacy standard. Argentina, the first non-European country to receive EU adequacy, has remained a leader in data protection reform.

A discussion on neuroscience and privacy, moderated by Ella Mein, explored ethical concerns surrounding brain data. Professor Marcello Ienca warned of potential ‘neuro-discrimination’ and the dangers of exploiting brain data. ECtHR Jurisconsult Anna Austin highlighted legal challenges, noting the high standards required for data waivers.

AD 4nXcD9vY3hv9j7oCdb0G0J3djLV7N9V owSFFxepiiG13uQenhtaLJAW0BQ0qTBaXJAbPEDoIusJTELxEi7NSYVqmAc87RTfHHhUgFc uamsGLzwD2VtcqroGc6UcrU5

The final panel, led by Gabriela Zanfir-Fortuna, addressed the need for stronger enforcement. Johnny Ryan of the Irish Council for Civil Liberties criticised the EU’s slow response to data misuse, while Nora Ni Loideain emphasised the GDPR’s role in giving DPAs greater enforcement power.

The event underscored the need for robust regulation, global cooperation, and better enforcement mechanisms to protect privacy in a rapidly evolving digital landscape.


Legacy media vs social media and alternative media channels

The rapid spread of digital information has transformed communication, offering opportunities and challenges. While social media and alternative platforms have democratised access to information, they have also enabled misinformation, deepfakes, and sensationalism to flourish. The tension between traditional media and these new forms of communication is at the heart of current debates on content policy and media integrity.

The case of Novak Djokovic at the 2025 Australian Open highlights this shift. After a Channel 9 journalist made derogatory remarks about him and his Serbian supporters, Djokovic refused an on-court interview and took to social media to share his perspective. His video went viral, attracting support from figures like Elon Musk, who criticised traditional media as a ‘negativity filter,’ This incident underscored the ability of social media to bypass mainstream media narratives, raising questions about journalistic objectivity, editorial oversight, and the role of direct communication in shaping public discourse.

AD 4nXehrAKIU3EadoVpFZcHaGsCRqgp5YUIbGGiFSd2bEOuE8Nbm7BqgDyG a35 XmdduVCs8BwGEoTbjQuJJmdZqfnwkza2wQmvAXSmigRGS8d7FtowkdzvZ2qMeGYb

Similarly, alternative media’s influence on political discourse was evident in Joe Rogan’s podcast, The Joe Rogan Experience. In 2024, Donald Trump’s appearance on the podcast allowed him to engage with audiences outside traditional news constraints, potentially boosting his presidential campaign. In contrast, Kamala Harris declined an invitation after requesting pre-approved questions. This difference illustrated how politicians navigate new media environments—some embracing unscripted discussions, others preferring controlled narratives. The case of Vladimir Klitschko further demonstrated how alternative media offers international figures a platform for nuanced discussions on global issues.

AD 4nXdYmGquEOcZTUg6pPP0En GQUhtXGWvKtuSBlfeE8 dsTMTOQo8yjLWFlvCN0G9I SPwW8mnfzyTj5czy 0QI72W7z XeDyXDwZfAx35khVizd0HUFL98NDtUPTxbSZdi2Q

Elon Musk’s experience with the media further highlights these dynamics. After traditional media misrepresented a gesture he made at a public event, Musk turned to X (formerly Twitter) to counter the narrative. His criticism of ‘legacy media’ as biassed and slow to adapt resonated with many, reinforcing the growing preference for direct, unfiltered communication. However, this shift presents risks, as social media and alternative platforms lack the editorial oversight that traditional outlets provide, allowing misinformation to spread more easily.

The rise of podcasts, independent media, and social networks has disrupted traditional journalism, offering new ways to engage audiences. While these platforms provide greater freedom of expression, they raise concerns about accuracy, misinformation, and accountability. The challenge remains in balancing openness with factual integrity, ensuring that media—whether traditional or alternative—serves the public responsibly in an era where the boundaries between truth and fabrication continue to blur.


New Cambridge school to explore AI’s economic impact

The University of Cambridge is set to open a new department, The Bennett School of Public Policy, later this year, dedicated to exploring how AI can improve both business and public sector operations.

However, this marks the university’s first major new academic department in the 21st century, with research priorities focused on harnessing AI to address pressing policy challenges.

Professors from a wide range of disciplines, including economics, social sciences, and medical fields, will collaborate to investigate the potential benefits and risks of AI adoption across sectors.

The school’s leadership aims to foster a generation of ‘tech-savvy’ policymakers who can effectively manage AI’s role in society while promoting sustainable and inclusive growth.

In addition to investigating AI’s economic impact, researchers will work closely with the Civil Service to develop AI workflows and help local government leaders address geographical inequality.

With a focus on cross-disciplinary expertise, the Bennett School plans to train policymakers who can tackle the complex issues of today’s world.

For more information on these topics, visit diplomacy.edu.

DeepSeek’s transparency in AI profit projections challenges industry norms

Chinese AI start-up DeepSeek has revealed that its theoretical profit margins could reach 545 percent, shedding light on the financial potential of its innovative approach to AI. The Hangzhou-based company said its V3 and R1 models’ cost-to-sales ratio during 24 hours in February highlighted the massive profitability of its AI services.

However, DeepSeek clarified that actual revenues are much lower, as many services remain free or discounted during off-peak hours, and the calculations exclude research and development costs. Despite the hypothetical nature of these margins, the disclosure comes as AI start-ups worldwide struggle to prove their business models can turn sustainable profits.

Giants like OpenAI and Anthropic experiment with various revenue streams — from subscriptions to licensing fees — while investors increasingly question how soon these companies can achieve lasting profitability. What makes DeepSeek stand out is its commitment to transparency and open-source innovation.

The company publicly shared insights into its infrastructure, explaining how it optimises server loads, reduces latency, and maximises data processing efficiency. This openness contrasts sharply with many US-based competitors’ secretive, proprietary strategies, signalling DeepSeek’s confidence in its long-term vision.

Why does it matter?

In an industry racing to build more powerful AI, DeepSeek’s willingness to share its methods and potential profitability sparks an important debate: Can a balance between innovation, openness, and sustainable revenue reshape the future of AI?

For more information on these topics, visit diplomacy.edu.

Samsung unveils new Galaxy A-series models with AI features and enhanced photography capabilities

Samsung has unveiled its latest Galaxy A-series lineup, introducing the Galaxy A56, A36, and A26, packed with new AI-powered features. One of the standout additions is the ‘awesome intelligence’ suite, which brings advanced image editing capabilities.

Among these features is Best Face, allowing users to swap facial expressions in group photos — similar to Google Pixel’s Best Take. These tools, first seen in the Galaxy S25, aim to simplify capturing and editing moments.

The Galaxy A56 has performance upgrades, including an Exynos 1580 chip and a larger vapour chamber for better cooling. It features a 6.7-inch full HD Plus display with a 120Hz refresh rate, a 50MP primary camera, a 12MP ultrawide sensor, and a 5MP macro lens. Priced at $499, the A56 is set to launch in the US later this year, while the A36 and A26 will hit shelves in late March, costing $399 and $299, respectively.

Buyers in the UK can get their hands on the new devices starting 19 March, with prices beginning at €299 for the Galaxy A26. The series also promises long-term value, offering up to six years of Android OS and security updates, plus IP67 dust and water resistance on the A26, making these phones a compelling choice for budget-conscious consumers looking for modern features.

For more information on these topics, visit diplomacy.edu.

Zhipu AI secures over 1 billion yuan in new funding

Chinese AI startup Zhipu AI has raised over 1 billion yuan ($137.22 million) in new funding, following a previous 3 billion yuan investment in December. The latest funding round comes amid growing competition in China’s AI sector, particularly with rival DeepSeek’s large language models, which claim to rival Western models at a lower cost.

Investors in the new round include the state-backed Hangzhou City Investment Group Industrial Fund and Shangcheng Capital. Zhipu AI plans to use the funds to enhance its GLM language model and expand its AI ecosystem, with a focus on businesses in Zhejiang province and the Yangtze River Delta region.

Founded in 2019, Zhipu AI has completed 16 funding rounds and is considered one of China’s leading AI startups. The company aims to release a range of new AI models, including foundation and multimodal models, as part of its open-source strategy. The funding comes as competitors like DeepSeek continue to disrupt the AI landscape with open-source models that challenge established platforms.

For more information on these topics, visit diplomacy.edu.

Singapore examines fraud case involving AI chip transfers

Singapore‘s government is investigating a fraud case involving servers supplied by US companies, potentially containing Nvidia’s advanced chips. Three men, including a Chinese national, were charged last week in connection with the alleged illegal transfer of these AI chips from Singapore to Chinese firm DeepSeek. Singapore’s Home Affairs and Law Minister K Shanmugam confirmed that the servers were provided by Dell Technologies and Super Micro Computer to local firms before being sent to Malaysia.

Authorities are still unsure whether Malaysia was the final destination, but they are working with US officials to determine if the servers contained restricted US export-controlled items. The US is already investigating whether DeepSeek has used banned Nvidia chips, which could lead to violations of export laws. The case forms part of a broader probe into suspected smuggling activities linked to AI chips being moved from countries like Singapore to China.

Singapore, a key market for Nvidia, is also examining allegations that DeepSeek may have acquired thousands of advanced Nvidia chips illegally. However, DeepSeek has denied these claims, stating that it only used legally purchased chips, including the Nvidia H800 model. The investigation is ongoing, with Singapore continuing to cooperate with US authorities.

For more information on these topics, visit diplomacy.edu.

Honor to expand into AI-powered devices with $10 billion plan

Chinese smartphone maker Honor plans to invest $10 billion over the next five years to develop AI for its devices as it prepares for a public listing. CEO James Li revealed at the Mobile World Congress in Barcelona that the company aims to expand beyond smartphones into AI-powered PCs, tablets, and wearables.

Honor’s push into AI comes after completing a shareholder restructuring in December, bringing it closer to an initial public offering, though no timeline has been set. The announcement also aligns with a surge in AI investment in China, driven by the popularity of DeepSeek’s affordable language models.

While Honor’s market share in China slipped from second to fourth place last year, it has gained significant support from the Shenzhen government, including R&D funding and tax breaks. The company is also focusing on expanding internationally as it strengthens its AI capabilities.

For more information on these topics, visit diplomacy.edu.

Senator Warren urges stricter AI chip controls on China

Senator Elizabeth Warren has called for tougher restrictions on Chinese technology, urging President Donald Trump’s nominee for a key trade role to reinforce AI chip controls.

In a letter to Jeffrey Kessler, the nominee for the Commerce Department’s Bureau of Industry and Security, Warren cited concerns over Chinese startup DeepSeek and its AI advancements. She argued for tighter enforcement of export restrictions and measures to curb chip smuggling.

Chinese firms like Huawei and SMIC, already on the US Entity List, continue to acquire American technology through front companies, Warren warned.

She pressed Kessler to address this issue and determine whether these companies violated US laws by producing advanced chips using American technology. Strengthened controls on ChangXin Memory Technology and high-powered inference chips, such as Nvidia’s H20, were also among her recommendations.

Kessler, who previously served as the Commerce Department’s top trade enforcement official, acknowledged China‘s progress in cutting-edge technology. His confirmation hearing before the Senate Banking Committee is set to examine how the US can maintain its technological edge while enforcing export restrictions effectively.

The debate comes amid ongoing tensions over AI and semiconductor technology between Washington and Beijing.

For more information on these topics, visit diplomacy.edu.

OpenAI launches GPT-4.5 for Pro users

OpenAI has rolled out a research preview of its latest GPT-4.5 model for Pro users, with plans to extend access to a wider audience over the next few weeks. The upgraded model boasts better pattern recognition, enhanced emotional intelligence, and an improved ability to generate creative insights.

The new version of GPT-4.5 also addresses a common issue with AI models, ‘hallucinations’ or the generation of false information, reducing its rate to 37.1%, compared to previous models’ higher rates. OpenAI’s CEO Sam Altman acknowledged some logistical challenges in the rollout, including GPU shortages, but assured users that more resources would be added soon.

While GPT-4.5 can handle tasks such as coding and content creation, it currently lacks some advanced features, such as voice and video capabilities. As OpenAI expands its offering, the company continues to refine the model to meet the growing demand for more powerful AI tools.

For more information on these topics, visit diplomacy.edu.

Cyber threats in 2024 shift to AI-driven attacks and cloud exploits, says CrowdStrike

A new report from CrowdStrike, the US-based cybersecurity company, examines the evolution of cyber threats in 2024, identifying shifts toward malware-free intrusions, artificial intelligence-assisted social engineering, and cloud-related vulnerabilities.

The researchers highlight an increase in cyber activity attributed to state-linked actors, a rise in identity-based attacks, and the growing role of generative AI in cyber operations. According to the report, 79% of cyber intrusions in 2024 did not involve traditional malware, compared to 40% in 2019. Attackers increasingly relied on remote management and monitoring tools to evade security measures. The average breakout time—the time taken for an attacker to move laterally within a compromised network—decreased to 48 minutes, with some intrusions occurring in under one minute.

The report also highlights an increased reliance on exploiting vulnerabilities, particularly for initial access. More than 52% of vulnerabilities observed in 2024 were related to gaining an initial foothold in a system, underscoring the importance of securing entry points. Attackers increasingly leveraged chained vulnerability exploits—where multiple flaws are exploited in succession—to enhance their chances of success.

Cloud security incidents also saw an increase, with valid account abuse accounting for 35% of cloud-related intrusions. Attackers focused on services such as Microsoft 365 and SharePoint, as well as enterprise APIs, to gain unauthorized access and extract data. The report emphasizes that more than half of observed vulnerabilities in 2024 were related to initial access, with an increase in attacks using chained vulnerability exploits.

Generative AI played a growing role in cyber operations, including phishing, deepfake-based social engineering, and automated disinformation campaigns. The report cites activity from groups leveraging AI-powered tactics, such as the use of fake job interviews to infiltrate technology firms.

For more information on these topics, visit diplomacy.edu.