The theme of the 2026 OTT Conference was Think Tanks and Trust; a topic that can easily become abstract and limit the conversation to jargon: credibility, legitimacy, influence, reputation, independence, transparency and impact as if these were institutional assets that can be easily designed, measured and managed.
But the four keynotes of the conference suggested something that was simultaneously deeper and more practical. Trust is not simply something think tanks possess. It is something they make and remake through relationships, judgment, humility, listening, storytelling, disagreement and responsibility.
In different ways, Rosa Balfour, Erica Schoder, Héctor Cárdenas and Denis Foretia each brought us back to the “soft” dimensions of think tank work. Soft, however, should not be confused with secondary. These are not decorative additions to research quality or policy influence. They are increasingly at the centre of what makes think tanks useful, credible and necessary.
Beyond evidence alone
For a long time, think tanks have justified themselves through the quality of their evidence.
But the conference keynotes pushed us to recognise that evidence alone is no longer enough — if it ever was.
Rosa Balfour began with a question that many in the sector will recognise: Do think tanks still matter? Instead of simply deafening expertise at all costs, Rosa located think tanks in a world where expertise itself has been destabilised by a range of factors: democratic backsliding, political polarisation, disinformation, geopolitical upheaval, technological disruption and declining trust in institutions. In this context, think tanks matter not merely because they produce analysis but because they can serve as bridges between policymakers and society at a time when both are under pressure.
This bridging role is not technocratic. It requires a moral clarity that is often absent from think tank discussions. This brought us back to Ruth Levine’s 2020 keynote on the ethical dilemmas for think tanks.
Rosa argued that think tanks need to know who they are, what they stand for and what kinds of work genuinely align with their mission. In a fragmented environment, authenticity becomes part of credibility. Organisations that chase every funding opportunity or react constantly to the crisis of the day risk losing the very identity that makes others trust them. Wasn’t this the challenge that Sara Pantuliano sought to address when ODI Global embarked on its recent business model refresh?
This was a useful reminder that trust does not begin with communication. It begins with purpose.
Trust as encounter
Erica Schoder’s keynote brought this question into the age of artificial intelligence. If AI can produce summaries, reports, briefs, literature reviews and analysis at extraordinary speed, what remains distinctive about think tanks?
AI may become one of the most powerful research partners that think tanks have ever had. But Erica argued that the real value of think tanks lies not only in their final output. It is also found in the process through which judgment is formed: the conversations, disagreements, editorial debates, stakeholder engagements and collective attempts to make sense of difficult problems.
AI may make some of the invisible work of think tanks more visible. If a report can be generated quickly, then we are forced to ask: what was the value of the work around the report? What judgment was built? What relationships were strengthened? What forms of trust were created? What did junior researchers learn? What did stakeholders help us see differently?
Erica called attention to the risk of “thinner knowledge”: outputs that look sophisticated but are detached from the encounters that produce understanding. This is particularly important for think tanks because their work is rarely just about information; it is about interpretation, judgment, and responsibility.
In this sense, trust is generated through encounters.
Trust as storytelling
Héctor Cárdenas took us to another human dimension of think tank work: storytelling. He argued that evidence does not travel alone. It travels through stories.
That old adage that “evidence speaks for itself” was finally laid to rest.
Héctor’s idea of “trustworthy storytelling” was grounded in engagement (echoing Erica), truthfulness, humility and political realism. Think tanks must show up where people are, study the problems that affect people, communicate in ways they can understand, acknowledge uncertainty, and root their narratives in lived experience and political context.
Think tanks cannot continue to assume that expertise breeds authority and that authority generates attention. As Héctor put it, too often we do not have audiences; we have artefacts. Reports, briefs exist and events are produced and reported on, but do people engage? Do they understand? Do they know where to go deeper? Do they trust think tanks enough to follow the argument beyond the headline of social media posts?
His keynote also warned against confusing visibility with trust. Short-form content, podcasts, AI-generated formats and social media can all help think tanks reach new audiences and accumulate followers and likes. But having people’s attention is not the same as being seen by them as credible. This is built through follow-through: the encounters that Erica argued for.
But those encounters will not sustain themselves. They need stories.
This is an important distinction for think tanks. We cannot complain that less evidence-based actors dominate public debate if we refuse to tell better stories.
Trust as local legitimacy
Denis Foretia’s keynote asked another key question: trusted by whom?
This question is especially important for think tanks working in contexts where international recognition, donor funding, and external validation can substitute for local legitimacy. Denis argued that trust is not one relationship. It is a series of relationships — with governments, funders and citizens — that can pull organisations in different directions.
A think tank may be respected by donors but unknown to its own citizens. It may be well known internationally, but marginal in its own policy community. It may produce research that satisfies foreign reporting frameworks but does not change the lives of the people it claims to serve. That, Denis warned, is not legitimacy. It is performance.
His reflections on “borrowed legitimacy” were particularly powerful. Too many think tanks rely on standards of recognition that come from elsewhere: international rankings, global visibility, donor approval, elite networks or external validation. These may be useful, but they are not the same as being trusted by the communities and institutions among whom we work.
It reminded me of Cameron Neylon’s 2019 keynote on excellence in which he argued that, too often, organisations rely on external – and not relevant – sources of legitimacy.
Denis spoke to the Nkafu Policy Institute’s experience with informal market traders in Cameroon. The organisation had supported formalisation as a route to credit, protection and business support. Technically, the idea made sense. But the project failed because it did not fully understand how the traders saw the problem. Their reluctance to formalise was not ignorance. It was mistrust of the state. Formalisation exposed them to unpredictable taxation, arbitrary payments and vulnerability to officials. The technically correct solution was misread socially and politically.
In Héctor’s words, the story lacked humility and political feasibility!
Before designing solutions, listen. Trust cannot be claimed on behalf of citizens. It has to be earned from them.
Four people-centred practices
Taken together, the keynotes suggested that think tanks need to pay much greater attention to four people-centred practices.
The first is purpose. Rosa reminded us that think tanks need moral clarity. Without it, independence becomes a slogan, strategy becomes opportunism, and credibility becomes fragile.
The second is encounter. Erica showed that knowledge is not only produced by experts working alone, but through relationships, disagreement and shared problem-solving. The room is bigger than we think.
The third is storytelling. Héctor argued that ideas matter only when they are heard, understood and able to survive contact with politics. Evidence needs a narrative, but the narrative must remain truthful.
The fourth is listening. Denis challenged us to ask who trusts us, why, and on what terms. Local legitimacy cannot be borrowed. It must be built.
These practices are not soft in the sense of being vague or abstract. They are hard because they require institutional discipline, capabilities and skills.
They ask think tanks to make choices: about funding, governance, communications, research design, partnerships, audiences and accountability.
They also require humility. A think tank that wants to be trusted must be willing to ask whether it has misunderstood the problem, whether its recommendations are usable, whether its audiences recognise themselves in its work, whether its partnerships are clear, whether its claims to independence are credible, and whether its knowledge is being produced with others rather than merely about them.
The future of trust
The conference took place at a moment when think tanks face many structural pressures: AI, polarisation, shrinking civic space, funding uncertainty, geopolitical instability and declining trust in expertise. But the four keynotes suggested that the future of think tanks will not be secured only by better tools, stronger brands or more sophisticated outputs.
It will depend on whether think tanks can deepen the human foundations of their work.
Can they create spaces where people with different perspectives can think together? Can they tell stories without distorting evidence? Can they listen before prescribing? Can they remain independent without becoming detached? Can they be useful without becoming instrumental? Can they be visible without becoming performative? Can they use AI without hollowing out the processes through which judgment is formed?
To echo our speakers, these are not just communications questions. They are institutional questions.
Perhaps the most important message I take away from the four keynotes is that trust is not built at the end of the research process, once the report is ready and the communications team begins its work.
Trust is built throughout: in the choice of questions, in the methods used, in the people invited into the process, in putting together the research team, in how disagreements are managed, in the stories told, in how uncertainty is acknowledged, in the relationships maintained throughout and after the process, and in the courage to remain truthful when it is inconvenient for the project or the organisations.