Artificial intelligence (AI) is changing how knowledge is produced, communicated, and consumed. For organisations such as think tanks, whose value rests on turning evidence into ideas that influence policy, the implications are transformative. Findings from the 2026 State of the Sector survey suggest that while AI adoption is widespread, deeper organisational integration remains a key challenge.
The Survey reveals that the think tank community is embracing the technology: 84 per cent of surveyed organisations report using AI in their work, while only 11 per cent say they do not.
This mirrors the broader trend of AI use expanding rapidly across organisations. McKinsey’s 2025 global survey found that 88 per cent of respondents reported regular AI use in at least one business function, up from 78 per cent a year earlier. Yet widespread adoption had not translated into deep organisational integration: most organisations remained in the experimenting or piloting stages, and only 7 per cent reported that AI has been fully scaled.
The distinction between adoption and integration matters for think tanks, as using AI to augment a narrow range of workflows captures only part of its potential.
How AI has entered the think tank toolkit
The State of the Sector survey finds that think tanks use AI where generative tools have obvious, immediate applications. Among the 251 organisations using AI, 76 per cent use it for communications and 75 per cent for research. 82 per cent of adopters use grammar and writing tools, 71 per cent use note-taking tools, and 58 per cent use general-purpose chatbots. As knowledge workers, think tank researchers spend considerable time searching for information, reviewing literature, transcribing interviews, and drafting documents — all tasks AI can shorten.
The immediate productivity gains are appealing. AI can help researchers create, manage, and examine large datasets; produce literature summaries; and complete tedious research tasks more quickly. Communications teams can use it to adapt technical research for different audiences, and administrative teams can automate repetitive processes. Used effectively, AI frees staff to spend more time on the work where think tanks add the most value: asking the right questions, engaging policymakers, and translating complex evidence into arguments that shape policy.
AI integration has further to go
However, the survey shows that AI use remains concentrated in particular functions. Among adopters, just 26 per cent apply AI to fundraising, a third of the rate at which they apply it to communications and research. Yet fundraising is the function think tanks most consistently identify as the one they need to strengthen.
Alongside its research applications, AI could help identify prospective funders, monitor funding opportunities, analyse donor priorities, and maintain donor databases. Its potential extends to knowledge management and internal administration — the functions where small organisations lose the most time.
Many think tanks, then, remain in the first phase of adoption, using readily available tools to make individual tasks easier rather than asking how AI might improve organisational performance more broadly.
This pattern is not unique to think tanks. Deloitte’s 2026 State of AI in the Enterprise found that workforce access to approved AI tools increased by 50 per cent in a single year, from fewer than 40 per cent to around 60 per cent of workers. Yet only 34 per cent of surveyed companies reported using AI to significantly transform their business models. The parallel suggests that the next challenge for think tanks is moving from individual experimentation to institutional strategy.
The value of AI still depends on people
Deeper integration also carries risks, and they fall hardest on organisations whose standing rests on credibility. A poorly sourced statistic, fabricated reference, or misleading interpretation can damage a reputation built over years. The speed AI offers cannot come at the expense of research integrity.
AI systems can produce inaccurate information, reproduce biases embedded in their training data, and raise concerns around privacy and intellectual property. UNESCO’s guidance on generative AI in education and research addresses these risks, asking institutions to keep decisions in human hands, protect the data they hold, and ensure a tool is fit for purpose before adopting it.
The State of the Sector findings show these concerns are already influencing AI adoption. Among the relatively small group of 33 organisations not using AI, 58 per cent identify ethical concerns as a barrier, and 45 per cent cite a lack of expertise. This highlights that access to AI is not the same as the capacity to use it well.
Therefore, think tanks need AI literacy alongside AI adoption. Researchers need to understand when AI is useful, when its outputs require verification, and when it should not be used at all. Organisations need clear rules governing confidential information, attribution, fact-checking, and disclosure of AI-assisted work.
From efficiency to capability
Smaller think tanks are poised to gain more from AI. Tasks that once required substantial staff time can increasingly be supported by AI, allowing organisations to expand their research, communications, and operational capacity. While this will not eliminate resource disparities, it can narrow the gap between what smaller teams want to achieve and what they can deliver.
The consequential gap will be between think tanks that pursue AI integration strategically and those that lag behind. As AI augments what think tanks can do, expectations of the speed, quality, and range of their outputs will also rise. In this context, AI integration is becoming more than a productivity issue; it is existential. Think tanks that fail to adapt risk becoming less competitive, less influential, and less relevant as sector expectations continue to rise. OECD research suggests that weak infrastructure, skills shortages, and adoption costs may make this challenge greater in lower-income settings.
For think tanks and funders, investment must extend beyond purchasing tools to building staff capabilities, institutional policies, and responsible-use frameworks. The survey shows that adoption is widespread. The next challenge is deeper integration: using AI strategically to strengthen research, organisational resilience, and policy influence while preserving the human judgement on which think tank credibility depends.