Impact capital in emerging economies increasingly seeks market-rate returns, leaving early-stage ventures that require system-building without suitable financing. Traditional blended finance models are faltering as impact investing converges with conventional commercial investing, favoring debt instruments with larger ticket sizes, which exclude smaller, riskier ventures. This shift disproportionately affects sectors like agri-SMEs in Sub-Saharan Africa, where financing gaps reach $65 billion annually. Concurrently, development assistance decreases, while many ventures, often infrastructure-heavy, are designed for rapid returns rather than building systemic trust and coordination. Closing this gap requires tailored capital instruments aligned with venture stages and a system-focused approach to growth.
Africa | Financial services
How to close the financing gap for emerging-economy ventures
Pakistan’s rapidly growing population exposes the limitations of its current administrative system, which struggles to serve vastly uneven and large provincial populations, such as Punjab’s 128 million versus Balochistan’s fewer than 15 million. Decentralization efforts, like the 18th Constitutional Amendment, have devolved responsibilities and resources to provinces, yet local decision-making remains distant from many communities, worsening service delivery gaps, especially in education. Effective reform requires aligning boundary changes with fiscal capacity, clear mandates, and accountability. Lessons from countries like Kenya show that creating smaller, well-funded, and politically empowered local units can improve governance, while fragmented or underfunded units, as seen in Lebanon, hinder service delivery. Pakistan’s reform options range from moderate to ambitious, balancing governance efficiency with local responsiveness.
Africa, Latin America | Health, Financial services
What Pakistan reveals about administrative reform for fast-growing populations
Leadership in the age of AI demands balancing machine-generated insights with human judgment, accountability, and oversight. As AI systems grow more capable in analysis, decision-making, and task execution, leaders must define clear boundaries for AI use, ensuring responsible governance and assessing real-world consequences. AI literacy becomes essential not only for technologists but also for policy-makers and executives to question AI outputs effectively. Preparing future leaders involves fostering adaptability, problem-finding, human-centric soft skills, and intellectual agility. The UAE's National Experts Program exemplifies practical AI leadership development by integrating AI knowledge with sector-specific challenges and national priorities.
Education
What does leadership in the age of AI look like in practice?
Research on Earth's climate tipping points—critical thresholds leading to abrupt changes in systems like ice sheets, rainforests, and ocean currents—has grown unevenly. The Greenland Ice Sheet dominates the literature, while others like abrupt boreal permafrost thaw or the North Atlantic subpolar gyre are far less studied, despite having similarly low warming thresholds (~1.5–1.8°C). Only a small fraction (8.1%) of studies focus on risk-relevant dynamics such as abruptness and irreversibility. This uneven attention highlights gaps in understanding important tipping behaviors, suggesting the need for increased research in less-studied areas to better inform climate risk assessments before these tipping points are reached.
Antarctica | Health
Some climate tipping points are being studied far more than others – new research
Urban noise from traffic, trains, and nightlife poses significant health risks, including disturbed sleep, stress, and increased heart disease. In Europe, noise is the second largest environmental health risk after air pollution, with about one in five urban residents exposed to harmful levels. Denmark and England illustrate contrasting approaches: Denmark treats noise as both a public health issue and planning concern, setting strict thresholds that limit development in noisy areas and emphasizing noise reduction measures like lower speed limits and barriers. England lacks national noise thresholds, allowing local authorities to balance noise with housing and economic priorities, resulting in greater variability in residents' noise exposure. Urban noise is not inevitable but can be managed through deliberate city planning.
How much noise should we put up with? London and Copenhagen have different answers
Autonomous AI systems like humanoid robots, self-driving vehicles, and drones are merging software and physical risks, dissolving traditional lines between cybersecurity and safety. These AI systems perceive, decide, and act in the physical world, making errors or attacks potentially harmful before human intervention. Humanoid robots in industry and healthcare, autonomous vehicles in road transport, and drones in delivery and military use illustrate advancing applications but also highlight governance challenges including safety certification, remote connectivity vulnerabilities, traffic coordination, and airspace management. Effective governance must integrate security and safety from design through deployment, clearly defining operational boundaries and human responsibility to manage this convergence of cyber and physical risks.
ICT
Why cybersecurity and physical autonomous AI must be governed together
Global investment in computing infrastructure surged to $459 billion in 2025, projected to increase tenfold by 2027 alongside advances in AI, quantum computing, and other technologies. The Next-Generation Computing Framework, developed by the World Economic Forum and its Global Future Council, addresses challenges of securing computing power amid energy demands, geopolitical tensions, and cyberthreats. It focuses on three pillars—sovereignty and interdependence, cybersecurity and resilience, and sustainability—across seven dimensions. The framework’s scorecard aids governments and organizations in assessing capabilities, setting priorities, and making informed investment and policy decisions to ensure computing advancements are secure, sustainable, and inclusive globally.
ICT
The Next-Generation Computing Framework
Global investment in computing infrastructure reached $459 billion in 2025 and is projected to increase tenfold by 2027. Advances in artificial intelligence, quantum computing, and other technologies present challenges in securing computing power while managing energy consumption, geopolitical dependencies, and cyber threats. The Next-Generation Computing Framework, developed by the World Economic Forum and its Global Future Council, offers a practical tool for governments and organizations. It focuses on three pillars—sovereignty and interdependence, cybersecurity and resilience, and sustainability—across seven dimensions, providing a scorecard to assess capabilities, set priorities, and guide investment and policy decisions toward secure, sustainable, and inclusive computing.
ICT
The Next-Generation Computing Framework
El Niño, currently at record strength and expected to peak around November, poses immediate risks such as failed rains, heatwaves, and higher food prices. However, its most severe economic impact is likely in 2027, as El Niño's effects can persistently slow productivity growth, especially in hot, dry economies like many in Asia and the Pacific. Agriculture-dependent regions face prolonged damage, with poor rural households and women most affected. Recovery is complicated by debt, global conflicts, and market strains. Early government action is crucial, including prearranged financing, protecting productive assets, activating social protection based on forecasts, managing inflation carefully, and coordinating comprehensive response plans to minimize long-term harm.
Africa, ASEAN, Latin America
Asian Development Blog: El Niño's Long Shadow: Why 2027 Is the Year to Worry About
Fertility rates have declined globally over the past 30 years, but this decline has not consistently aligned with increased female labor participation. Despite urbanization and economic growth, women's engagement in paid work remains uneven and has stagnated or decreased even in some upper-middle-income countries. The fertility–labor nexus, which examines the relationship between fertility outcomes and women's labor market participation, varies across countries and income levels. Patterns differ notably among Organisation for Economic Co-operation and Development member countries, highlighting the importance of gender equality and social policies in shaping these dynamics.
Africa, Latin America | Health
Reassessing the Fertility–Labor Nexus: Why Gender Equality and Social Policy Matter
AI's integration in workplaces demands careful delegation of tasks based on human versus AI strengths. Humans excel in authentic relationship-building, empathy, moral accountability, and adaptability—areas where AI falls short. The HUMAN framework offers five criteria for leaders: Humanizing (requiring a genuine human touch), Unscripted (needing spontaneous judgment), Moral (decisions requiring human accountability), Adaptive (adjusting beliefs based on new perspectives), and maintaining trust. While AI can handle data-driven tasks, roles involving empathy, improvisation, and responsibility are best suited for humans.
Education
Is This a Job for a Human or AI?
Indigenous communities seek not only representation in AI but also the ability to shape, govern, and benefit from it, especially amid AI’s growing influence in education, public services, and culture. AI offers potential for Indigenous language revitalization, cultural transmission, and economic opportunities, but risks perpetuating data extraction and misinformation without inclusive data governance. Addressing this requires new approaches to data commons, infrastructure, and investment to ensure AI respects Indigenous values and knowledge, preventing harm and enabling community empowerment.
Education
Data Commons as Infrastructure for Indigenous-Led AI
Democracy remains a valid concept but is hindered by outdated practices rooted in the 19th century, leading to declining trust, disengagement, and participation. Geoff Mulgan and Alex Fischer propose viewing democracy as an interconnected system of functions—including elections, decision-making, public services, and citizen participation—that can be enhanced. They emphasize leveraging digital technology and AI to build new cognitive infrastructures that improve collective intelligence, decision-making, and trust. Their framework promotes ongoing democratic innovation and renewal rather than mere preservation.
Africa | Health, Education, ICT
Work in Progress: How to Upgrade Democracy in the Age of AI
Citizen science is expanding worldwide due to advancements in technology and increased public involvement in research. Volunteers assist scientists by collecting vital data over large geographic areas and extended periods. In tourism, participants engage in activities like coral planting, wildlife tracking, and environmental DNA collection. Crowdsourced data helps monitor climate change, watersheds, and disaster preparedness. Despite challenges such as data quality, reliability, bias, and institutional hesitation, combining volunteer efforts with professional expertise effectively bridges data gaps in environmental research.
ICT, Health
Why Citizen Science Is On The Rise
The next generation of data centres will be defined by resilience, focusing on climate risk, engineering, and community priorities. The global surge in digital infrastructure investment, driven by AI and digitalization, demands data centres that withstand long-term climate challenges. Site selection increasingly incorporates climate risk alongside power and connectivity, as 17% of US data centre capacity faces high natural catastrophe risks, especially in states like Texas. Developers use climate analytics to assess evolving risks, influencing design and financing. Engineering decisions must address rising temperatures and water scarcity, balancing operational efficiency with environmental strain. Resilience planning now includes both acute disasters and gradual climate shifts.
3 key factors that will define the next generation of data centres
The UK state pension is less generous than most of Europe’s, replacing 54% of average earnings including workplace pensions, below the OECD average of 63%. The state pension alone replaces just 22% compared to the OECD public scheme average of 43%. Despite rising costs partly due to the triple lock system, demographic factors mainly drive increased spending. The UK spends less on public pensions (7.1% of GDP) than the OECD average (8.1%), relying more on private saving. Reforms announced by PM Andy Burnham will end earnings-linked increases from 2030, aiming to save £15 billion annually by the late 2030s, funding a National Care Service, though debate on costs and impacts continues.
Britain’s state pension is less generous than most of Europe’s
Anthropic's AI system Claude identified a novel family of enzymes called array-associated reverse transcriptases (ARTs) in bacteriophages, viruses that infect bacteria. This discovery highlights AI's potential role in biological research. However, computational biologist Mario Rodríguez Mestre questioned whether Claude's finding was truly independent or influenced by unpublished research he shared with the AI. Anthropic denies using user data for model training or that its team accessed such data. The enzymes, related to reverse transcriptases that convert RNA into DNA, were previously identified, but Anthropic characterized a larger system involving repeated DNA sequences producing small RNAs. While promising, the discovery requires further validation and peer review. Ethical AI use in research demands careful data handling, transparency, and adherence to institutional guidelines.
A biologist queries whether AI ‘borrowed’ his work to make major discovery – an expert explains the rules
Hospital appointment timing significantly impacts patients' travel modes and ease, especially for older or unwell individuals. Early appointments can precede concessionary bus pass validity, and specialist hospital visits often require complex travel, causing long, exhausting journeys. Transport challenges like unreliable buses, long distances, and multiple connections contribute to missed appointments and generate over 40% of NHS transport emissions. Studies highlight that travel burdens peak around diagnosis, particularly for cancer patients. Solutions include patient-centred appointment scheduling aligned with public transport, coordinating multiple appointments, remote consultations, and providing local services. Enhancing public transport accessibility and offering transport support like vouchers can improve healthcare access while reducing emissions.
Health
Hospital appointments can make greener travel harder
Tech companies collect extensive personal data not only to improve their products but also to target advertising and sell information to data brokers, sometimes including sensitive details about individuals’ vulnerabilities. Smartphones may not be actively listening to conversations, but firms use various inferred data, raising trust concerns. Sharing personal data affects not just individuals but also their families and communities. Although data collection began as an accidental byproduct of digital services, it has become a central business model, particularly for companies like Google, which monetize user access rather than just data. Alternative payment models beyond data-sharing are possible and worth pursuing in democratic societies.
ICT
Is your phone listening to you? “We must be more mindful of the kinds of companies we engage with”
Smartphones listen only for wake-up words like "Alexa" or "Siri" and discard other sounds, contrary to the common belief they constantly eavesdrop. Social media ads seemingly related to private conversations often result from data on friends' interests, not from active listening. Algorithms track user engagement—pausing, liking, and watching videos—to personalize content, creating "filter bubbles" that reinforce preferences. While personalized ads serve business goals and user relevance, the same mechanisms can skew news exposure and perspectives, especially among youth relying on social media for information. Data collection is generally legal with user consent, though consent processes are often not fully understood or thoughtfully given. Improved regulation should address genuine user choice and consent transparency.
Is your phone listening to you? Or do social media firms just know what to serve you?