EU watchdog urges limits on US data access

The European Union’s data protection watchdog has urged stronger safeguards as negotiations continue with the US over access to biometric databases. European Data Protection Supervisor Wojciech Wiewiórowski said limits must ensure Europeans’ data is used only for agreed purposes.

Talks between the EU and the US involve potential arrangements that would allow US authorities to query national biometric systems. Databases across the EU contain sensitive information, including fingerprints and facial recognition data.

Past transatlantic data-sharing agreements between the two have faced legal challenges due to insufficient safeguards. European regulators are closely monitoring the Data Privacy Framework amid ongoing concerns about oversight.

Officials also warned that emerging AI technologies could create new surveillance risks linked to US data access. European authorities said they must negotiate as a unified bloc when dealing with the US.

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

Major crypto exchanges in South Korea face new ownership limits

South Korea’s ruling Democratic Party and the Financial Services Commission have agreed to cap major shareholder stakes in domestic crypto exchanges at 20%. Exceptions of up to 34% would apply to new businesses to support early-stage operators.

Large exchanges like Upbit and Bithumb will have 3 years to comply, while smaller platforms will receive an additional 3-year grace period.

Current ownership exceeds the proposed cap, with Upbit at 25.5%, Bithumb at 73.6%, and Coinone at 53.4%. Korbit’s pending acquisition would give Mirae Asset Consulting 92% ownership, highlighting the extent of concentrated holdings in the market.

The cap seeks to curb governance risks from concentrated shareholding, following the FSC’s January 2026 proposal. The move gained urgency after Bithumb’s accidental $43 billion Bitcoin transfer, which raised concerns about internal controls.

The ownership limit will likely be included in South Korea’s upcoming Digital Asset Basic Act, alongside rules on stablecoins and crypto ETFs.

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

Passkey login comes to Windows 11 via Bitwarden vault

Bitwarden has announced support for logging into Windows 11 devices using passkeys stored in its encrypted vault, enabling phishing-resistant authentication directly at the operating system login screen.

The feature is available across all Bitwarden plans, including the free tier, and is believed to be a first for a third-party password manager.

During the login process, Windows 11 displays a QR code that users scan with their mobile device running the Bitwarden app, which then confirms access to the stored passkey and completes authentication.

Unlike device-bound passkey implementations, passkeys are synchronised across devices via Bitwarden’s end-to-end encrypted vault, meaning users can still regain access even if their phone is lost.

The feature builds on Microsoft’s introduction of native support for external passkey managers in Windows 11 in November 2025. It requires the device to be joined to Microsoft Entra ID with FIDO2 security key sign-in enabled.

Microsoft says the passkey-based login will roll out throughout March, depending on an organisation’s Entra ID configuration.

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

TikTok rejects end-to-end encryption citing safety concerns

TikTok will not adopt end-to-end encryption for direct messages. The company explained that using this technology could hinder safety teams’ and law enforcement’s efforts to detect harmful content in private messages, which the company believes could make users less safe online.

Encrypted messaging ensures that only the sender and recipient can read a conversation and is widely used across the social media industry. Rivals including Facebook, Instagram, Messenger, and X have adopted the technology, saying protecting private communication is central to user privacy.

The issue has become more sensitive because the platform has long faced scrutiny over possible links between its parent company, ByteDance, and the government of the People’s Republic of China, something the company has repeatedly denied. Reflecting these concerns, earlier this year, US lawmakers ordered the separation of TikTok’s US operations from its global business.

The company told the BBC that encrypted messaging would make it impossible for police and platform safety teams to read direct messages when needed. TikTok emphasised that this decision was made to enhance user protection, with a particular focus on the safety of younger users, and that it sees monitoring capabilities as crucial for addressing harmful behaviour.

Industry analyst Matt Navarra said the platform’s decision to ‘swim against the tide’ is ‘notable’ but presents ‘challenging optics’. He noted, ‘Grooming and harassment risks are present in DMs [direct messages], so TikTok can state it is prioritising proactive safety over privacy absolutism,’ though he added that the decision ‘places TikTok out of alignment with global privacy expectations’.

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

Online privacy faces new pressures in the age of social media

Online privacy is eroding as digital services collect ever-growing personal data and surveillance becomes part of daily technology use. The debate has intensified as social media platforms, advertisers, and connected devices expand their ability to track behaviour, preferences, and habits.

Analysts say younger generations have adapted to this reality rather than resisting it. ‘In 2026, online privacy is a luxury, not a right,’ says Thomas Bunting, an analyst at the UK innovation think tank Nesta. He argues many people have grown up accepting data collection as a trade-off for access to online services, noting: ‘We’ve been taught how to deal with it.’

Advocates warn that the erosion of online privacy could have wider social consequences. Cybersecurity expert Prof Alan Woodward from the University of Surrey says the issue goes beyond personal privacy. ‘People should care about online privacy because it shapes who has power over their lives,’ he says, arguing that privacy is ‘about having something to protect: freedom of thought, experimentation, dissent and personal development without permanent surveillance.’

Despite a growing number of privacy tools and regulations, data exposure remains widespread. According to Statista, more than 1.35 billion people were affected by data breaches, hacks, or exposure in 2024 alone. At the same time, more than 160 countries now have privacy legislation, while users regularly encounter cookie consent prompts that govern how their data is collected online.

Experts say frustration with privacy controls reflects a broader ‘privacy paradox’, in which people express concern about data protection but rarely change their behaviour. Cisco’s Consumer Privacy Survey found that while 89% of respondents said they care about privacy, only 38% actively take steps to protect their data.

As philosopher Carissa Véliz notes, the challenge is not simply awareness but a sense of agency: ‘Mostly, people don’t feel like they have control.’ She argues that protecting privacy requires stronger regulation, responsible technology design, and cultural change, adding: ‘It’s about having [access to] the right tech, but also using it.’

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

UK to launch new lab for breakthrough AI research

Researchers in the UK will gain a new AI lab designed to drive transformational breakthroughs in healthcare, transport, science, and everyday technology, supported by government funding.

The lab will provide up to £40 million in funding over six years, alongside substantial access to large-scale computing resources, inviting UK researchers to pitch their most ambitious ideas.

The Fundamental AI Research Lab will focus on tackling core AI challenges, including hallucinations, unreliable memory, and unpredictable reasoning.

The lab will support high-risk, blue-sky research rather than simply scaling existing systems. Its goal is to unlock entirely new capabilities that could improve medical diagnoses, infrastructure resilience, scientific discovery, and public services.

UK officials highlighted the country’s strength in world-class universities, AI talent, and a thriving sector attracting over £100 billion in private investment. Experts, including Raia Hadsell of Google DeepMind, will peer-review funding applications, prioritising bold, high-reward proposals.

The initiative is part of the UKRI AI Strategy, which is backed by £1.6 billion and aims to strengthen research and ensure AI benefits society and the economy. UK AI projects like RADAR for rail faults and the IXI Brain Atlas for Alzheimer’s research demonstrate the approach’s potential impact.

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

AI adoption and jobs debated at India summit

Governments, companies and international organisations gathered in India in February for the AI Impact Summit to discuss the future of AI governance and adoption. Participants in India focused on economic impacts, labour market changes and sector specific uses of AI.

Delegates in India also highlighted growing interest in international cooperation on AI governance. Ninety one countries endorsed a declaration supporting shared tools, global collaboration and people centred development of AI.

Language diversity became a central topic during discussions in India. India’s government announced eight foundation AI models designed to support generative AI across the country’s 22 recognised languages.

Debate in India also reflected the growing influence of the Global South in AI policy discussions. Policymakers and experts in India emphasised infrastructure gaps, language diversity and local economic realities shaping AI adoption.

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

ECB reports minor impact of AI on employment

AI has so far had only a small effect on employment across Europe, according to economists at the European Central Bank. A comparison of 5,000 firms- both AI users and non-users- showed no significant difference in job creation or reduction.

Some firms that use AI intensively were even four percent more likely to hire new staff than average.

Economists noted that AI investment has not replaced existing jobs. In some cases, firms are hiring additional employees to develop and implement AI systems or to scale up operations more efficiently.

Only a minority of firms, around 15 percent, reported reducing labour costs as a motivation for AI adoption.

Despite limited impacts so far, the ECB cautioned that AI could have more significant effects as technology matures. Firms that specifically invest in AI to cut jobs may indeed reduce employment, and the long-term consequences for production processes and labour markets remain uncertain.

The findings come amid rising concern over AI-driven job losses, with companies such as Amazon and Allianz citing AI as a reason for recent cuts. Markets reacted negatively last week after a viral post predicted widespread layoffs, though current evidence shows only minor effects.

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

OpenAI tracks how AI shapes student performance over time

AI is increasingly shaping education, offering tools like ChatGPT that provide personalised learning support for students anywhere. Early studies suggest features such as study mode can enhance exam performance, yet understanding AI’s long-term effect on learning remains a challenge.

Traditional research often focuses on test scores, overlooking how students interact with AI over time in real-world settings.

OpenAI, in partnership with Estonia’s University of Tartu and Stanford’s SCALE Initiative, created the Learning Outcomes Measurement Suite to track longitudinal learning outcomes. The framework assesses interactions, engagement, cognitive growth, and alignment with pedagogical principles.

Large-scale trials involve tens of thousands of students, combining AI-driven insights with traditional classroom measures such as exams and observations.

Research shows that guided AI interactions can strengthen understanding, persistence, and problem-solving. Microeconomics students using the study mode achieved around 15% higher exam scores than those relying on traditional online resources.

Beyond short-term results, the measurement suite evaluates deeper learning effects, including motivation, metacognition, and productive engagement, helping educators and developers optimise AI tools for meaningful outcomes.

The suite will be validated through ongoing studies and eventually made available to schools, universities, and education systems worldwide. OpenAI aims to share findings broadly to ensure AI contributes effectively to student learning and cognitive development.

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

Council of Europe issues new guidance on AI and gender equality

Ahead of International Women’s Day on 8 March, the Council of Europe adopted two new recommendations addressing gender equality and the prevention of violence against women in the context of emerging technologies.

One recommendation targets the design and use of AI to prevent discrimination, while the other focuses on accountability for technology-facilitated violence against women and girls.

The AI recommendation advises member states on preventing discrimination throughout the lifecycle of AI systems, from development to deployment and retirement. It highlights risks like gender bias while promoting transparency, explainability, and safeguards.

Special attention is given to discrimination based on gender, race, and sexual orientation, gender identity, and expression (SOGIESC).

The second recommendation sets the first international standard for addressing technology-facilitated violence against women. It outlines strategies to overcome impunity, including clearer legal frameworks, accessible reporting systems, and victim-centred approaches.

Emphasis is placed on multistakeholder engagement, trauma-informed policies, and safety-by-design in technology products to prevent digital harm.

Both recommendations reinforce the importance of combining regulation, institutional support, and public awareness to ensure technology advances equality rather than perpetuates harm.

The formal launch is scheduled for 10 June 2026 at the Palais de l’Europe in Strasbourg during an event titled ‘From standards to action: making accountability for technology-facilitated violence against women and girls a reality.’

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