As the International Science Council meets in Beijing, and China publishes the first open, system-level account of its own fundamental research, James Wilsdon asks whether AI is expanding the science system or diluting it – and how the ISC should respond.

Trust in science is not collapsing, but it is being fragmented, politicised and renegotiated. Lise Korsten argues that this reflects friction between an evolving scientific method and a digital information ecosystem that gives evidence and misinformation similar visibility. Internal integrity problems and the alignment of scientific views with political identity add to the strain. The era of unquestioned expertise is over. To renew its social contract, the scientific community needs to show humility about what is not yet known, teach the public how science works, and protect inquiry from political and corporate agendas.

As the largest technological powers pursue strategic autonomy, countries without the scale to build their own technological ecosystems risk being left with strategic dependence. Nanditha Mathew and Luc Soete argue that fragmentation is creating an “access gap”, cutting off the knowledge, infrastructure and networks through which capabilities are built, with consequences for development and global inequality. For these countries, they say, full technological sovereignty is neither feasible nor the right goal. The aim should be technological agency: keeping meaningful choices over how technologies are accessed, governed, adapted and used.

As AI and other digital technologies become strategic assets, governments are recasting the purpose of science around national competitiveness and security. Nanditha Mathew and Luc Soete argue that this “nationalisation of scientific purpose”, visible in both the US and Europe, erodes the institutional autonomy, openness and international collaboration on which scientific leadership depends. The challenge is not to insulate science from national priorities, but to stop the pursuit of strategic advantage from undermining the openness that makes scientific progress possible.

Tackling complex societal challenges takes interdisciplinary solutions, yet these remain hard to achieve. Prof Ruth Morgan argues that, for the most intractable problems, progress depends less on new knowledge than on synthesising the specialist knowledge that already exists. That means seeking out the contradictions between disciplines, not just the common ground, and using them to build new shared language and ask new questions. She calls on science institutions to create spaces for ongoing dialogue, foster a culture of challenge, and value the communicators, translators and mediators who make this work possible.

Decades of science advice have not stopped the planet heading towards preventable catastrophe, and scientific institutions are failing to deliver on their mission. Dr Tracey Elliott argues that “neutral” science advice must give way to responsible advocacy and activism. Research cultures should recognise and reward it, especially for early- and mid-career researchers.

As private firms now dominate R&D in most OECD economies, Wagner, Soete, and Mathew argue this isn’t just a funding shift — it’s rewiring how scientific knowledge is disclosed, shared, and governed, with corporate secrecy replacing open publication in the very sectors (like AI) that matter most. They call for international policy action to protect the knowledge commons before the system’s asymmetries harden into place.

As populism, AI, and political transactionalism reshape how governments use expertise, Sir Peter Gluckman argues that science advice must evolve from slow, analogue evidence synthesis to fast, AI-assisted brokerage — with human experts still essential to frame questions, curate outputs, and keep policy makers honest about what science does and doesn’t know.

Dr Michael S. Ringel, Life Biosciences, USA; Dr Carolyn B. Ringel, Harvard Medical School, USA — Biomarkers of aging can inform drug trials and clinical decisions as supportive evidence while formal validation continues. 

Legal infrastructure, not just science, decides whether precision biomedicine reaches patients. This piece argues data governance must be built into the system from the start — open where safe, controlled where necessary, accountable throughout the full data-to-model lifecycle.

When AI pre-digests contested evidence before advisers see it, errors can compress in and disagreement can compress out. Laura V. Zimmermann argues that a flawless-looking AI briefing on labour trafficking still buried a data error and hid real scientific dispute — and asks what advisers and institutions must do as fewer, shared models increasingly shape what counts as evidence.

Dr Ricardo Baptista Leite, HealthAI – The Global Agency for Responsible AI in Health, Switzerland — Health systems must be redesigned around prevention before advances in medical AI can reduce disease burden and narrow inequities in care.

Science diplomacy has shifted amid growing geopolitical disruption, pushing scientists and diplomats toward more aspirational, pragmatic, and transactional approaches. In response, Québec announced at the 2025 UNESCO Global Ministerial Dialogue that it would build a network of research chairs in science diplomacy — a plan realized in March 2026 when the FRQ launched eight chairs covering themes from AI governance to Arctic diplomacy and food sustainability. This marks a shift from ad hoc engagement to structured policy capacity, now formalized in Québec’s international policy since June 2026.

Lanthanide probes and associated AI should be assessed together within a unified international framework, argue Profs Pingkun Yan and Ge Wang, involved in IEEE projects related to medical imaging, and Prof Mannudeep K. Kalra, radiologist.

As geopolitical rivalry strains the institutions that have underpinned global cooperation since 1945, José Manuel Durão Barroso argues that science can still hold the door open. In this piece, the former European Commission President makes the case for science diplomacy as a bridge across political divides, for open collaboration balanced with legitimate security concerns, and for treating knowledge as Europe’s “Fifth Freedom” — a strategic asset as vital to the continent’s future as the free movement of goods, people, and capital.

The glucose ketone index (GKI) and other evidence-based biomarkers have an important role in personalized nutrition care in the management of chronic diseases. However, realizing their potential requires policy changes to ensure equitable access to personalized nutrition care via competency-based education and licensure and reimbursement reform—argue Corinne Bush and Amy Smith of the American Nutrition Association. 

Facing mounting geostrategic, political, and technological pressures, science cannot afford to be passive—it must confront the incentives and cultural failings within its own institutions, from the “bibliometric disease” to weakened self-regulation. Only by pre-emptively reforming how it governs, communicates, and justifies itself can science shape its own future and remain a trusted public good, argues Sir Peter Gluckman.

Science has entered a data-rich era, but its integrity, equity, and openness now depend on treating FAIR data as a first-class output and closing the governance gaps around AI-driven research. Without urgent policy action on data standards, infrastructure, and incentives, AI risks accelerating hallucination and eroding trust rather than advancing trustworthy science, argue Barend Mons, Arie Baak, and Koen Jonkers.

In a rapidly shifting geopolitical landscape, scientific institutions must proactively reform their internal incentives and self-governance to protect research integrity and ensure knowledge remains a trusted global public good. Passive compliance with external pressures will no longer suffice if science is to survive as a credible, universally accessible enterprise, argues Sir Peter Gluckman.

Prof Rita Carsetti, International Union of Immunological Societies, Germany, and Prof Michael Schmueck-Henneresse, University Health Network, Canada — Bringing tolerance-restoring Treg therapies to patients worldwide requires regulation, manufacturing, and reimbursement to evolve alongside the science.