AI in Strategic Foresight 2025

The insights from the OECD-World Economic Forum survey highlight a pivotal moment for the foresight profession. While the findings confirm AI’s potential to enhance efficiency, they also reveal a range of challenges and uncertainties. Opportunities Increasing AI literacy: A notable finding is the disparity in perceived AI skills across sectors, with public sector respondents reporting a more modest assessment of their capabilities compared to the private sector. This may necessitate a coordinated and systematic increase of AI literacy in the public sector through training and active experimentation. Practitioners, particularly those without prior experience, often have limited ideas about AI’s potential, and can therefore be less well positioned to identify, and avoid, its pitfalls. An active, experimental approach is crucial, as the survey found that experience with AI significantly increases the perception of its usefulness. This is not simply about adopting off-the-shelf tools but also about automating different parts of the strategic foresight process, augmenting human creativity and developing tailored solutions. Moving beyond simple use cases: The survey indicates that most strategic foresight practitioners currently use AI for simpler tasks, such as synthesizing data, initial scanning and sense- making. While this is a valuable starting point, the future lies in more advanced applications. Practitioners should invest time in using AI as a sparring partner and idea generator, and to systematize and summarize signals, suggest scenarios and compare collected data. The aim is for a level of maturity where AI is integrated into the entire strategic foresight process, with tailored tools developed for complexity mapping, pattern detection and communication of findings. Developing human-centred workflows: The survey reinforces the notion that AI tools are purely supplemental and require careful analysis and existing human expertise. Strategic foresight practitioners should invest in developing workflows that leverage AI to handle the “heavy lifting” of data processing and initial drafts, thereby freeing up time for higher-level analysis, interpretation and critical thinking. The objective is to use AI to enable capabilities previously infeasible, such as automated signal detection and large-scale document analysis, while maintaining human oversight for nuance and originality in addition to verification of outputs.Challenges The survey highlights that while most practitioners are optimistic about AI’s potential, they also recognize significant risks. Guarding against unreliable and biased outputs: The most frequently cited challenge is the quality and trustworthiness of AI- generated outputs. This includes the concern of hallucinations, weak and shallow content, and a general lack of originality or imagination. Practitioners must maintain a critical mindset and implement robust verification protocols to fact-check AI outputs, especially since a lack of transparency regarding sources and logic is a major concern. Different AI tools could be used in parallel to validate and check results. Addressing ethical and governance gaps: The survey found that a significant challenge is a lack of clear ethical guidelines and governance. This is compounded by data security and confidentiality restrictions, which prevent the feeding of sensitive internal documents into AI engines. Strategic foresight practitioners may need to advocate for and help develop ethical frameworks that address issues of data ownership, accountability and responsible use of AI.8 Across countries, there is a lack of guidance and resources to experiment with AI in government in a responsible way.9 Tackling skill gaps: The survey highlighted a differing AI literacy rate among strategic foresight experts. Successfully experimenting with and integrating AI into strategic foresight processes also requires a certain level of skills in both AI and data management. Targeted hiring and internal upskilling and training programmes for strategic foresight teams can help increase the internal capacity needed to ensure a more systemic uptake of AI. Overcoming organizational inertia: The survey showed that respondents in the public sector, civil society and academia face resistance from leadership or other stakeholders when integrating AI. This is linked to a general climate of risk aversion and a lack of resources and time for the necessary experimentation. Practitioners need to build a compelling case for the value of AI, demonstrating its benefits through small-scale experiments and pilots. Successfully integrating AI into foresight may be one of the best ways to demonstrate the systemically connected issues affecting any organization and, as such, this integration could help governments overcome exactly those silos.Conclusion AI in Strategic Foresight: Reshaping Anticipatory Governance 18