Future of work

EU decision regulates researcher access to data under the DSA

A document released by the Republican-led House Judiciary Committee revived claims that the EU digital rules amount to censorship. The document concerns a €120 million fine against X under the Digital Services Act and was framed as a ‘secret censorship ruling’, despite publication requirements.

The document provides insight into how the European Commission interprets Article 40 of the DSA, which governs researcher access to platform data. The rule requires huge online platforms to grant qualified researchers access to publicly accessible data needed to study systemic risks in the EU.

Investigators found that X failed to comply with Article 40.12, in force since 2023 and covering public data access. The Commission said X applied restrictive eligibility rules, delayed reviews, imposed tight quotas, and blocked independent researcher access, including scraping.

The decision confirms platforms cannot price access to restrict research, deny access based on affiliation or location, or ban scraping by contract. The European Commission also rejected X’s narrow reading of ‘systemic risk’, allowing broader research contexts.

The ruling also highlights weak internal processes and limited staffing for handling access requests. X must submit an action plan by mid-April 2026, with the decision expected to shape future enforcement of researcher access across major platforms.

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

AI governance becomes urgent for mortgage lenders

Mortgage lenders face growing pressure to govern AI as regulatory uncertainty persists across the United States. States and federal authorities continue to contest oversight, but accountability for how AI is used in underwriting, servicing, marketing, and fraud detection already rests with lenders.

Effective AI risk management requires more than policy statements. Mortgage lenders need operational governance that inventories AI tools, documents training data, and assigns accountability for outcomes, including bias monitoring and escalation when AI affects borrower eligibility, pricing, or disclosures.

Vendor risk has become a central exposure. Many technology contracts predate AI scrutiny and lack provisions on audit rights, explainability, and data controls, leaving lenders responsible when third-party models fail regulatory tests or transparency expectations.

Leading US mortgage lenders are using staged deployments, starting with lower-risk use cases such as document processing and fraud detection, while maintaining human oversight for high-impact decisions. Incremental rollouts generate performance and fairness evidence that regulators increasingly expect.

Regulatory pressure is rising as states advance AI rules and federal authorities signal the development of national standards. Even as boundaries are debated, lenders remain accountable, making early governance and disciplined scaling essential.

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

AI anxiety strains the modern workforce

Mounting anxiety is reshaping the modern workplace as AI alters job expectations and career paths. Pew Research indicates more than a third of employees believe AI could harm their prospects, fuelling tension across teams.

Younger workers feel particular strain, with 92% of Gen Z saying it is vital to speak openly about mental health at work. Communicators and managers must now deliver reassurance while coping with their own pressure.

Leadership expert Anna Liotta points to generational intelligence as a practical way to reduce friction and improve trust. She highlights how tailored communication can reduce misunderstanding and conflict.

Her latest research connects neuroscience, including the role of the vagus nerve, with practical workplace strategies. By combining emotional regulation with thoughtful messaging, she suggests that organisations can calm anxiety and build more resilient teams.

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

Half of xAI’s founding team has now left the company

Departures from Elon Musk’s AI startup xAI have reached a symbolic milestone, with two more co-founders announcing exits within days of each other. Yuhuai Tony Wu and Jimmy Ba both confirmed their decisions publicly, marking a turning point for the company’s leadership.

Losses now total six out of the original 12 founding members, signalling significant turnover in less than three years. Several prominent researchers had already moved on to competitors, launched new ventures, or stepped away for personal reasons.

Timing coincides with major developments, including SpaceX’s acquisition of xAI and preparations for a potential public listing. Financial opportunities and intense demand for AI expertise are encouraging senior talent to pursue independent projects or new roles.

Challenges surrounding the Grok chatbot, including technical issues and controversy over its harmful content, have added internal pressure. Growing competition from OpenAI and Anthropic means retaining skilled researchers will be vital to sustaining investor confidence and future growth.

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

Hybrid AI could reshape robotics and defence

Investors and researchers are increasingly arguing that the future of AI lies beyond large language models. In London and across Europe, startups are developing so-called world models designed to simulate physical reality rather than simply predict text.

Unlike LLMs, which rely on static datasets, world models aim to build internal representations of cause and effect. Advocates say these systems are better suited to autonomous vehicles, robotics, defence and industrial simulation.

London based Stanhope AI is among companies pursuing this approach, claiming its systems learn by inference and continuously update their internal maps. The company is reportedly working with European governments and aerospace firms on AI drone applications.

Supporters argue that safety and explainability must be embedded from the outset, particularly under frameworks such as the EU AI Act. Investors suggest that hybrid systems combining LLMs with physics aware models could unlock large commercial markets across Europe.

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

Researchers tackle LLM regression with on policy training

Researchers at MIT, the Improbable AI Lab and ETH Zurich have proposed a fine tuning method to address catastrophic forgetting in large language models. The issue often causes models to lose earlier skills when trained on new tasks.

The technique, called self distillation fine tuning, allows a model to act as both teacher and student during training. In Cambridge and Zurich experiments, the approach preserved prior capabilities while improving accuracy on new tasks.

Enterprise teams often manage separate model variants to prevent regression, increasing operational complexity. The researchers argue that their method could reduce fragmentation and support continual learning, useful for AI, within a single production model.

However, the method requires around 2.5 times more computing power than standard supervised fine tuning. Analysts note that real world deployment will depend on governance controls, training costs and suitability for regulated industries.

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

Latam-GPT signals new AI ambition in Latin America

Chile has introduced Latam-GPT to strengthen Latin America’s presence in global AI.

The project, developed by the National Centre for Artificial Intelligence with support across South America, aims to correct long-standing biases by training systems on the region’s own data instead of material drawn mainly from the US or Europe.

President Gabriel Boric said the model will help maintain cultural identity and allow the region to take a more active role in technological development.

Latam-GPT is not designed as a conversational tool but rather as a vast dataset that serves as the foundation for future applications. More than eight terabytes of information have been collected, mainly in Spanish and Portuguese, with plans to add indigenous languages as the project expands.

The first version has been trained on Amazon Web Services. At the same time, future work will run on a new supercomputer at the University of Tarapacá, supported by millions of dollars in regional funding.

The model reflects growing interest among countries outside the major AI hubs of the US, China and Europe in developing their own technology instead of relying on foreign systems.

Researchers in Chile argue that global models often include Latin American data in tiny proportions, which can limit accurate representation. Despite questions about resources and scale, supporters believe Latam-GPT can deliver practical benefits tailored to local needs.

Early adoption is already underway, with the Chilean firm Digevo preparing customer service tools based on the model.

These systems will operate in regional languages and recognise local expressions, offering a more natural experience than products trained on data from other parts of the world.

Developers say the approach could reduce bias and promote more inclusive AI across the continent.

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

European ombudsman opens probe into AI use in EU funding reviews

A formal inquiry has been opened into how AI is used in the evaluation of the EU funding proposals, marking the first investigation of its kind at the institutional level.

European Ombudsman Teresa Anjinho initiated the probe following allegations that external experts relied on AI systems when assessing applications.

Concerns emerged after a Polish company failed to secure support from the European Innovation Council Accelerator programme after submitting its bid before the November 2023 deadline. The complainant alleged that third-party AI use compromised fairness and influenced the assessment outcome.

Requests have been made for clearer governance standards, including explicit disclosure when AI systems are used in proposal reviews. Fears also emerged that sensitive commercial data could be exposed through external AI platforms.

Despite no grounds to reopen the case, a systemic probe into AI transparency and safeguards was launched. Document inspections are scheduled through March, followed by institutional meetings in April to determine whether regulatory or procedural changes are warranted.

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

AI safety leader quits Anthropic with global risk warning

A prominent AI safety researcher has resigned from Anthropic, issuing a stark warning about global technological and societal risks. Mrinank Sharma announced his departure in a public letter, citing concerns spanning AI development, bioweapons, and broader geopolitical instability.

Sharma led AI safeguards research, including model alignment, bioterrorism risks, and human-AI behavioural dynamics. Despite praising his tenure, he said ethical tensions and pressures hindered the pursuit of long-term safety priorities.

His exit comes amid wider turbulence across the AI sector. Another researcher recently left OpenAI, raising concerns over the integration of advertising into chatbot environments and the psychological implications of increasingly human-like AI interactions.

Anthropic, founded by former OpenAI staff, balances commercial AI deployment with safety and risk mitigation. Sharma plans to return to the UK to study poetry, stepping back from AI research amid global uncertainty.

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

Young voices seek critical approach to AI in classrooms

In Houston, more than 200 students from across the US gathered to discuss the future of AI in schools. The event, organised by the Close Up Foundation and Stanford University’s Deliberative Democracy Lab, brought together participants from 39 schools in 19 states.

Students debated whether AI tools such as ChatGPT and Gemini support or undermine learning. Many argued that schools are introducing powerful systems before pupils develop core critical thinking skills.

Participants did not call for a total ban or full embrace of AI. Instead, they urged schools to delay exposure for younger pupils and introduce clearer classroom policies that distinguish between support and substitution.

After returning to Honolulu, a student from ʻIolani School said Hawaiʻi schools should involve students directly in AI policy decisions. In Honolulu and beyond, he argued that structured dialogue can help schools balance innovation with cognitive development.

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