At the 2026 OTT Conference in Rabat, we asked whether think tanks can still be trusted.
By the end of the week, I found myself asking a slightly different question: what, exactly, is worth trusting them for?
This is not a semantic distinction. For years, think tanks have described their value through familiar words: evidence, independence, rigour, credibility, influence, and impact. These words still matter. But in Rabat, they felt insufficient. Across keynotes, workshops, public events, corridor conversations and drinks and dinners, a new vocabulary kept appearing: trust, care, respect, humility, legitimacy, encounter, storytelling, community, usefulness.
As an economist, I sometimes feel that I am wired to think of these as soft words. But experience tells me they are anything but soft. They may be the hardest words in the think tank lexicon.
They point to something I have been thinking about for some time: the future may not belong to think tanks as we have known them, but the think tank function remains essential. Organisations will change. Some will disappear. Some will become smaller, looser, more networked, more commercial, more embedded in universities, political parties, communities or digital platforms.
Some may even stop using the label altogether.
But the functions think tanks perform — making sense of public problems, convening disagreement, translating knowledge, preparing people for public life, holding open spaces for informed debate — are not going away.
If anything, they are becoming more important.
But they will have to be delivered differently.
Have we been looking for value in the wrong place?
For too long, think tanks have looked for their value in a limited set of outputs and outcomes.
A report published. A policy brief cited in a policy document. A minister briefed. An appearance in the media. A recommendation adopted. A conference was convened. A donor is satisfied.
These are not irrelevant. But they do not tell the whole story, not even its most important part.
In her keynote, Erica Schoder argued that artificial intelligence is making visible what has long been invisible in think tank work. If AI can produce a competent policy brief, then the value of a think tank cannot lie only in the policy brief.
If AI can be used by a minister to get up to speed with the latest evidence on their brief or quickly check the technical feasibility of a policy idea, then the value of a think tank cannot lie on whether they get cited in policy documents.
A think tank’s value must lie in the process that produces judgment: the disagreement over framing, the conversation with affected communities, the methodological argument, the editorial challenge, the ongoing process of helping frame and reframe the questions that policymakers ask, the trust built between people who see the world differently, etc.
In other words, the value is not just in the document or the citation but in and among the people that think tanks can bring together.
And the room that think tanks can fill is bigger than we think.
Anne-Marie Slaughter (New America) spoke of this 5 years ago when she put forward the change hub model for think tanks. The room should include thinktankers, but also the politicians, civil servants, journalists, funders, business leaders, activists, communities, critics, students and partners whose questions reshape the work. It also includes think tank communicators, project managers, fundraisers and planners. It also includes the junior researcher learning judgment through drafting and redrafting – or even listening in to the conversations of senior researchers while assisting in tasks that AI can now do faster (and better?).
And because not everyone thinks in the same way, the room must be host to the sometimes uncomfortable exchange that changes how a problem is understood, challenges an assumption and accepts that we do not know “what works.”
This is a hopeful future, though. The risk is that many conversations about the future of research, policy research, and think tanks are being reduced to a single question: AI adoption. As it was repeated throughout the Conference, AI is an extraordinary research partner, but it cannot take responsibility for meaning, it cannot (or should not) decide what matters, it certainly can’t build political trust, it cannot know what a community has lived through, and it cannot replace the slow formation of judgment in people and institutions.
The risk is not that AI will write the report. The risk is that because AI can write the report, we stop investing in the encounters that make it worth trusting.
Think tanks must become useful again
I have argued before that we should judge think tanks less by their supposed influence and more by their usefulness.
I think influence is often overstated. Policy change is rarely caused by one report, one meeting, one person, or one institution – no matter how large and full of brilliant minds. Think tanks may be present when decisions are made, but so are political incentives, public opinion, party dynamics, donors, media narratives, private interests, institutional constraints and – as one Conference participant so clearly put it – luck.
We should therefore be more modest about claiming influence.
Usefulness is a better standard.
Useful to whom? For what? Under what conditions? These are better questions than “Did we influence policy?”
Think tanks can be useful to governments trying to understand a problem, to opposition parties preparing for office, to journalists trying to interpret complexity, to civil society coalitions seeking arguments, to funders trying to understand an ecosystem, to young people looking for a path into public life, to communities trying to make sense of policies that affect them, to societies trying to preserve spaces for reasoned disagreement.
They can be useful for their own staff, giving them a sense of purpose.
I think the future lies in becoming more useful in contexts where evidence is needed, not just in producing more evidence!
But being useful requires, as John Schwartz put it, care; or as Anthea Haryoko did, respect; or as Denis Foretia and Héctor Cárdenas said, humility.
Rosa Balfour’s keynote asked whether think tanks still matter. Her answer was not a nostalgic defence of the sector but a challenge. Think tanks matter only if they can be useful to their societies at a time when trust in expertise, democratic institutions and shared facts is increasingly fragile.
But this role cannot be built with technical language alone. It requires moral clarity: a sense of who the organisation is, what it stands for and what work it should not do. It brought Robin Nibblet’s similar call to think tanks to discover their sense of purpose.
It requires self-doubt and critical thinking: the willingness to ask whether those who disagree with us may, in fact, have a point. A quality, I hope, OTT Conferences and the entire On Think Tanks programme have helped the sector to develop.
The think tanks that matter in the future will not be those that merely publish. They will be those who help others think.
Evidence does not travel alone
Héctor Cárdenas put this sharply in his keynote: evidence does not travel alone; it travels through stories.
This is difficult for many think tanks, which worry that storytelling means simplification, manipulation, branding, or spin. (Sometimes it does.) But the alternative is other people telling stories with weaker evidence, fewer scruples and greater emotional force.
If we do not tell the story, others will.
But Hector pointed out that our stories need to be better: truthful, grounded, humble and useful.
A better story can help people understand why an issue matters, connect technical analysis to lived experience, and show why a policy choice has consequences.
The future think tank will need to be both rigorous and narratively competent. It will need to know how to speak to different publics without becoming trapped by the logic of virality. Short formats may attract attention, but trust is built through follow-through. A video, podcast, or social media post may open the door, but there must be something deeper to go to and someone to talk to.
This has implications for brands, too. John Schwartz’s and Annapoorna Ravichander’s session on trusted think tank brands made clear that a brand is not a logo. It is the accumulation of thoughts and feelings people associate with an organisation. A trusted brand is an organisation people care about, and that cares back.
Care is strategic. Care means making knowledge usable, designing events that respect people’s time, publishing in formats that audiences can understand, listening before advising, not hiding behind jargon, and being visible for the right reasons, to the right people, at the right time.
If we think about this in this way, how many think tanks care?
Legitimacy cannot be borrowed forever
Denis Foretia warned that many think tanks enjoy borrowed legitimacy.
Cameron Neylon made a similar argument in his 2019 keynote speech.
Too often, think tanks seek legitimacy from sources that are foreign to their missions.
They are known internationally but not locally. They are visible and respected by other researchers and their peers, but not by citizens. They meet all donor quality and reporting requirements, but their business models are not rooted in the communities affected by the policies they study. They seek to be invited to global fora but may not be trusted at home.
This is not a sustainable foundation for the future.
For many think tanks, especially in the Global South, international funding and recognition have been both enabling and distorting. They have allowed institutions to emerge, hire staff, produce research and enter policy debates, but they have also encouraged business models and agendas that are not always grounded in local constituencies of support.
But when foreign funding disappears, what remains? A brand? A website? A set of reports? Or a local institution that people know, value and are willing to defend?
Or the opposite? An organisation with very smart people and great potential, but not known or valued locally, and a budget that no local funder is willing to cover.
This question is now urgent because funding cuts, shrinking civic space, restrictions on foreign funding and donor shifts towards security and defence are already reshaping the sector. The 2026 State of the Sector discussions during the Conference reinforced what many leaders know from experience: funding instability damages institutional memory, staff retention, long-term research agendas and trust.
The future of think tanks will depend partly on whether they can build more legitimate, locally grounded business models.
In lower and some middle-income countries, domestic funding matters more than ever, but it will not appear magically. In the session on mobilising domestic funding, examples from Brazil and Kenya showed that enabling environments for philanthropy require policy reform, coalitions, evidence and persistence. Think tanks can help build the conditions for their own sustainability: endowment laws, tax incentives, public benefit legislation, better philanthropy infrastructure and stronger domestic constituencies for evidence.
But they need to work together – and collaborate with other actors in academia and civil society – to deliver change.
The future is more local — and more global
One of the tensions running through the conference was between local legitimacy and global engagement.
Karim El Aynaoui’s welcoming remarks placed this tension at the centre of the conversation. Think tanks in the Global South are not simply local actors trying to catch up with global debates. They are producers of knowledge, narratives and talent that global debates urgently need. Rebalancing North–South narratives is not about symbolic inclusion but about who gets to define problems, whose evidence counts and whose futures are imagined.
The sessions on Global South engagement and regional think tanks made this concrete. Southern think tanks need to be present in global policy spaces because global decisions shape national policy options. But presence is not enough. The goal is knowledge sovereignty: being recognised as producers of original ideas, not merely providers of local data for Northern analysis.
To be “in the lead”, not just “in the loop”.
Regional think tanks play a particular role here. They can bridge national realities and global agendas, they can amplify smaller countries, and surface diversity within regions rather than flattening it into one voice.
Their advantage is in acting as marginal actors: they understand and operate local realities and global spaces at the same time.
But this role requires resources, time, language capacity, regional trust and agenda-setting power. Too often, regional work is funded through short-term projects designed elsewhere, into which Southern partners are invited as participants in bit roles but never as protagonists.
The future of think tanks will depend on whether regional and global cooperation becomes more genuinely plural. This applies to spaces such as the T7 and T20 processes, ASEAN think tank networks, APEC think tank networks and other similar spaces of engagement.
The question is not whether Northern institutions should include Southern voices. It is whether global policy conversations can be redesigned so that different centres of knowledge shape the agenda from the start.
The future of think tanks is in its people
Another theme that deserves more attention is talent.
Karim described PCNS as both a public-good institution and a talent factory. The student-led think tank session made the same point from a different angle. Think tanks are not only producers of knowledge but places where people learn how to participate in public life.
This has always been true, but it may become more important. If AI automates parts of research production, think tanks will need to ask how junior researchers will develop judgment. If funding becomes more unstable, they will need to ask how to retain talent during difficult months or years. If politics becomes more polarised, they will need to train people capable of working across disagreement. If public debate becomes more fragmented, think tanks will need communicators, facilitators, moderators, conveners, organisers, and political interpreters — not just researchers.
Student-led and youth-led think tanks already understand some of this. They are training grounds for future policy professionals. They teach writing, coding, research, leadership, event organisation, partnership-building and institutional memory.
Their challenge is continuity: students graduate, volunteers leave, and knowledge disappears. But their strengths are real: energy, experimentation, peer-to-peer learning, digital fluency and a willingness to build community.
Established think tanks should not treat student-led think tanks as junior versions of themselves. They should ask what they can learn from them.
The same applies internally. Organisational culture is no longer a “soft” concern. It is a strategic asset. Think tanks that cannot manage conflict, support staff, create psychological safety, offer progression or live their values under pressure will struggle to remain credible externally.
The future think tank will need to be designed around people: those it employs, those it trains, those it convenes and those it serves.
Think tanks as trust infrastructure
What, then, is the future of think tanks?
It is tempting to answer with organisational forms: more hybrid models, more university-linked centres, more networks, more digital-first organisations, more specialised boutiques, more individual experts with large audiences, more consultancy-like think tanks, more partisan institutions, more student-led or diaspora-led initiatives.
All of this may happen. The think tank community is already very diverse, and I do not expect it to lose this quality any time soon.
But the more important answer is functional.
The future of think tanks lies in becoming part of society’s trust infrastructure.
By trust infrastructure, I do not mean a certification scheme or a communications campaign, although such initiatives may have a role. I mean the practical systems, relationships, and habits that enable knowledge to be produced, tested, interpreted, and used responsibly.
Drawing from the Conference keynotes, sessions and conversations, I understand this trust infrastructure to include:
- Transparent funding.
- Financial models that do not distort think tanks’ missions.
- Clear, transparent and verifiable review processes.
- Responsible AI use.
- Spaces for disagreement.
- Multiple and overlapping communities of practice to address matters of shared interest.
- Local legitimacy.
- Strong and accountable governance.
- Public engagement first – not last.
- Ethical storytelling.
- Organisational cultures that allow people to speak honestly.
- Partnerships that are based on reciprocity rather than extraction and dependence.
- And more.
This infrastructure is not going to build itself, but the good news is that it is built through the work itself.
It is built when a think tank listens before designing a project. Or when it explains how it uses AI. It is built to make data understandable, not just available. It is built when it admits uncertainty and risk. Or when a think tank says no to funding that would compromise its mission. It is built when it gives young researchers real responsibility. Or when it creates a space where people who disagree can still think together.
The future of the function, not the label
At the conference, we had a brief discussion about the label: think tank. Should we defend its use? I do not think the future of think tanks will be secured by defending or ditching the label. The label has always been unstable, contested and context-specific. Some organisations will embrace it while others will reject it.
I think what really matters is whether the functions these organisations fulfil survive.
Can societies still find places where evidence is taken seriously but not worshipped? Where political realities are understood but not surrendered to? Where stories are told without abandoning truth? Where disagreement is structured rather than weaponised? Where young people learn the craft of public reasoning? Where local knowledge can shape global debates? Where can public institutions be challenged and helped at the same time?
If think tanks can be those places, they will matter.
They will matter because they are useful to their societies.
And in a world of abundant information, fragile institutions, polarised publics and increasingly powerful machines, usefulness may be the highest form of influence we can honestly claim.