Capacity development

Autonomous AI fails most tasks in virtual company experiment

Researchers at Carnegie Mellon University created a virtual company staffed solely by AI ’employees’ trained on large language models from vendors including Anthropic, OpenAI, and Google, assigning them roles such as financial analyst and software engineer.

In this simulated work environment, the AI agents struggled to complete most tasks, with even the best-performing model only completing about a quarter of its assignments.

The experiment highlighted key weaknesses in current AI systems, including difficulty interpreting nuanced instructions, managing web navigation with pop-ups, and coordinating multi-step workflows without human intervention.

These gaps suggest that human judgement, adaptability and collaboration remain essential in real workplaces for the foreseeable future.

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

UK firms prioritise cyber resilience and AI growth

Cybersecurity is set to receive the largest budget increases over the next 12 months, as organisations respond to rising geopolitical tensions and a surge in high-profile cyber-attacks, according to the KPMG Global Tech Report 2026.

More than half of UK firms plan to lift cybersecurity spending by over 10 percent, outpacing global averages and reflecting heightened concern over digital resilience.

AI and data analytics are also attracting substantial investment, with most organisations increasing budgets as they anticipate stronger returns by the end of 2026. Executives expect AI to shift from an efficiency tool to a core revenue driver, signalling a move toward large-scale deployment.

Despite strong investment momentum, scaling remains a major challenge. Fewer than one in 10 organisations report fully deployed AI or cybersecurity systems today, although around half expect to reach that stage within a year.

Structural barriers, fragmented ownership, and unclear accountability continue to slow execution, highlighting the complexity of translating strategy into operational impact.

Agentic AI is emerging as a central focus, with most organisations already embedding autonomous systems into workflows. Demand for specialist AI roles is rising, alongside closer collaboration to ensure secure deployment, governance, and continuous monitoring.

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

Face scans replace fingerprints in new motorcycle clearance trial at Singapore land border

The Immigration and Checkpoints Authority (ICA) has launched a facial recognition trial for motorcyclists entering Singapore via Woodlands Checkpoint, aiming to speed up cross-border clearance and improve convenience while maintaining security.

The pilot, launched on 26 January 2026, operates in two designated motorcycle lanes in the arrival zone and allows riders to use contactless facial scans rather than traditional fingerprint scans to verify identity.

Eligible users include Singapore residents, long-term pass holders and foreign visitors who have previously entered the country; no prior registration is needed to take part.

Riders simply scan their passport or MyICA QR code, lift their visor, and remove any obstructions (like sunglasses or masks) before looking into the facial recognition camera. ICA officers are on standby to assist and collect feedback to refine the system.

The initiative is part of ICA’s broader use of biometric technologies, including QR code clearance and iris/facial biometrics, to make immigration more efficient and contactless at Singapore’s land checkpoints.

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

Aquila transforms warehouse operations using AI automation

Aquila has completed a €5 million investment in AI-driven warehouse automation at its Ilfov, Dragomiresti logistics centre. The project is a strategic response to increasing portfolio complexity and growing distribution volumes in the FMCG sector.

The automation solution is built around AI-based vision systems that identify products directly from images using shape, colour and visual characteristics. The technology removes the need for labels or manual scanning, even when packaging orientation or appearance shows minor variations.

According to the company, the system improves the speed and accuracy of warehouse operations while reducing manual work and optimising storage space. These efficiency gains allow better use of operational resources.

The investment enables Aquila to scale logistics operations without proportional increases in resources. The company reports improved internal efficiency, stronger service quality for customers and the creation of medium-term competitive advantages.

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

The UK labour market feels a sharper impact from AI use

Companies are reporting net job losses linked to AI adoption, with research showing a sharper impact than in other major economies. A Morgan Stanley survey found that firms using the technology for at least a year cut more roles than they created, particularly across the UK labour market.

The study covered sectors including retail, real estate, transport, healthcare equipment and automotive manufacturing, showing an average productivity increase of 11.5% among UK businesses. Comparable firms in the United States reported similar efficiency gains but continued to expand employment overall.

Researchers pointed to higher operating costs and tax pressures as factors amplifying the employment impact in Britain. Unemployment has reached a four-year high, while increases in the minimum wage and employer national insurance contributions have tightened hiring across industries.

Public concern over AI-driven displacement is also rising, with more than a quarter of UK workers fearing their roles could disappear within five years, according to recruitment firm Randstad. Younger workers expressed the highest anxiety, while older generations showed greater confidence in adapting.

Political leaders warn that unmanaged AI-driven change could disrupt labour markets. London mayor Sadiq Khan said the technology may cut many white-collar jobs, calling for action to create replacement roles.

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

Maia 200 AI inference accelerator unveiled by Microsoft

Microsoft has unveiled Maia 200, a next-generation AI inference accelerator built to boost performance, efficiency, and cost-effectiveness at scale. Built on TSMC’s 3-nanometre process, the chip boosts speed, efficiency, and memory throughput for advanced AI models.

The new accelerator will power Microsoft’s cloud infrastructure across Azure, Microsoft Foundry, and Microsoft 365 Copilot, including workloads for OpenAI’s latest GPT-5.2 models.

Internal teams will use Maia 200 for synthetic data generation and reinforcement learning, accelerating AI development. Maia 200 is being rolled out in Microsoft’s US Central data centre region, with further deployments planned across additional global locations.

A preview version of the Maia software development kit is also being released, offering developers access to PyTorch integration, optimised compilers, and low-level programming tools to fine-tune AI models across heterogeneous computing environments.

The system introduces a redesigned networking and memory architecture optimised for high-bandwidth data movement and large-scale inference clusters.

Microsoft says the platform delivers significant improvements in performance per dollar, scalability, and power efficiency, positioning Maia 200 as a cornerstone of its long-term AI infrastructure strategy.

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

Siri set for major AI overhaul through Google’s Gemini partnership

Apple is preparing a major AI upgrade for Siri powered by Google’s Gemini models, expected in the second half of February, according to Bloomberg. The update will run on Apple’s Private Cloud Compute infrastructure using high-end Mac chips.

The iOS 26.4 release is set to introduce ‘World Knowledge Answers’, enabling Siri to provide web-based summaries with citations similar to ChatGPT and Perplexity. Deeper integration across core apps such as Mail, Photos, Music, TV, and Xcode is also planned.

Expanded voice controls are expected to let users search for and edit photos by spoken description, as well as generate emails based on calendar activity. Bloomberg also reported Apple is paying Google around $1 billion annually to access Gemini’s underlying AI technology.

Market reaction to the news pushed Apple shares higher, while Alphabet stock also rose following confirmation of the partnership. A spokesperson for Apple declined to comment on the reported developments.

Looking ahead, Apple is developing a chatbot-style assistant known internally as ‘Campos’ to eventually replace the current Siri interface. The system would analyse on-screen activity, suggest actions, and expand device control across future operating systems.

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

AI Overviews leans heavily on YouTube for health information

Google’s health-related search results increasingly draw on YouTube rather than hospitals, government agencies, or academic institutions, as new research reveals how AI Overviews select citation sources in automated results.

An analysis by SEO platform SE Ranking reviewed more than 50,000 German-language health queries and found AI Overviews appeared on over 82% of searches, making healthcare one of the most AI-influenced information categories on Google.

Across all cited sources, YouTube ranked first by a wide margin, accounting for more than 20,000 references and surpassing medical publishers, hospital websites, and public health authorities.

Academic journals and research institutions accounted for less than 1% of citations, while national and international government health bodies accounted for under 0.5%, highlighting a sharp imbalance in source authority.

Researchers warn that when platform-scale content outweighs evidence-based medical sources, the risk extends beyond misinformation to long-term erosion of trust in AI-powered search systems.

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

Google.org backs AI-led science breakthroughs

The organisation Google.org has selected twelve recipients for its $20 million AI for Science fund, which aims to accelerate research in health, agriculture, biodiversity, and climate.

The initiative backs academic, nonprofit, and startup teams using AI to turn scientific insights into real-world solutions. In health and life sciences, projects target genetic decoding, neural mapping, disease prediction, and faster detection of drug resistance.

Research groups are applying advanced AI models to unlock hidden regions of the human genome, simulate disease pathways, and dramatically reduce detection times for life-threatening pathogens, shifting medicine towards earlier intervention and prevention.

Agriculture and food systems are another focus, using AI to breed resistant crops, improve nutrition, and cut livestock methane emissions. Projects seek to strengthen food security, boost sustainability, and support climate resilience.

Biodiversity and clean energy efforts target species mapping, conservation planning, fusion research, and large-scale carbon capture. Open science principles ensure datasets and tools remain accessible, scalable, and capable of driving wider breakthroughs.

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

Indeed expands AI tools to reshape hiring

Indeed is expanding its use of AI to improve hiring efficiency, enhance candidate matching, and support recruiters, while keeping humans in control of final decisions.

The platform offers over 100 AI-powered features across job search, recruitment, and internal operations, supported by a long-term partnership with OpenAI.

Recent launches include Career Scout for job seekers and Talent Scout for employers, streamlining career guidance, sourcing, screening, and engagement.

Additional AI-powered tools introduced through Indeed Connect aim to improve candidate discovery and screening, helping companies move faster while broadening access to opportunities through skills-based matching.

AI adoption has accelerated internally, with over 80% of engineers using AI tools and two-thirds of staff saving up to 2 hours per week. Marketing, sales, and research teams are building custom AI agents to support creativity, personalised outreach, and strategic decision-making.

Responsible AI principles remain central to Indeed’s strategy, prioritising fairness, transparency, and human control in hiring. Early results show faster hiring, stronger candidate engagement, and improved outcomes in hard-to-fill roles, reinforcing confidence in AI-driven recruitment.

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