For example, multilateral bodies like my organisation, the Group on Earth Observations (GEO), have long focused on ensuring an open and trustworthy data ecosystem. But as private ventures increasingly outpace us, we must develop the capacity to innovate quickly, albeit transparently and inclusively.
Multilateral organisations like the GEO, United Nations agencies and development banks have struggled to keep up because they were not built for agility. Governance by many member states results in more bureaucracy and risk aversion, as well as funding that is tightly earmarked. Few multilateral initiatives contain sunset clauses to retire ineffective projects, while legacy structures and incentives make it difficult to discontinue old programmes and free up resources for new ideas. All this makes innovation painfully slow.
But these institutions can encourage experimentation and controlled risk-taking by adopting an accelerator model with which they can quickly test proposed solutions. As Strategyzer co-founder and CEO Alex Osterwalder has stressed, treating initiatives as a portfolio of bets, with some failures expected, lays the groundwork for breakthroughs. A corporate example is Bosch’s accelerator, which invested around US$30 million ($38.6 million) in 214 projects, treating each one as an investment with transformative potential.
This calculated risk-taking, which contrasts sharply with conservative multilateral financing models, ultimately led to 19 successful businesses.
See also: Trump, Xi hold call as trade, technology dispute roils ties
Systemic reforms
Encouragingly, some multilateral organisations have begun to embrace such a model. The UN Development Programme (UNDP) has created a huge development-innovation network, with 91 Accelerator Labs experimenting and adapting in order to improve outcomes for 115 countries. The UN World Food Programme has enhanced its impact with its Innovation Accelerator, which has mobilised nearly US$300 million since 2015 to deploy AI-powered tools and climate-resilience strategies. And GEO’s Global Ecosystems Atlas initiative went from concept to prototype in just nine months, demonstrating that rapid innovation is achievable with the right support.
The challenge now is to undertake systemic reforms that will make innovation a core competency of multilateral institutions. For starters, that means establishing accelerators, innovation labs, or venture teams within these organisations, and providing them with a clear mandate to seek and develop new ideas, as well as the autonomy to operate outside normal bureaucratic channels. For example, UNDP’s Accelerator Labs are safe spaces for experimentation, largely freed from political oversight while still maintaining transparency and clear governance.
Regular reviews
Multilateral institutions should also establish mechanisms to conduct regular reviews of programmes. In addition to defined milestones, every initiative should be phased out or terminated if it does not deliver results after a certain period, a common tech-industry practice that prevents stale projects from draining resources indefinitely. Adopting this portfolio approach would ensure a healthy pipeline of programmes: high-risk ideas enter the portfolio, and then are either scaled up (if successful) or set aside (if underperforming). As the UK’s funding ultimatum to underperforming agencies has demonstrated, ending or overhauling a failing program is sometimes necessary to make progress.
See also: Vietnam sends trade department reply as US turns up China heat
Equally important is earmarking funds for high-risk, high-reward experiments. This “venture fund” within a multilateral institution can be used to co-finance pilot projects or tech trials without diverting resources from core programmes. In addition to introducing risk capital, these organisations should adopt innovation accounting, which uses metrics focused on learning and early impact, such as the number of prototypes tested or pivots made. Inspired by lean startup methodology, these benchmarks hold teams accountable for validated learning and enable data-driven decisions about which pilots to scale or shut down. Instead of avoiding failure at all costs, multilateral institutions must learn how to fail productively.
Similarly, innovation will falter if staff are terrified that managers will punish their every misstep. Leaders must actively promote a risk-tolerant culture. That means celebrating teams that test bold ideas (even if the results are mixed), instituting rewards for creativity and problem-solving, and training teams in smart risk-taking. Organisations like the World Bank and UNDP have started to cultivate this kind of environment by sponsoring innovation challenges and recognising internal “intrapreneurs.”
Lastly, traditional hierarchies and lengthy planning cycles are at odds with the innovation process. Multilateral institutions should embrace more agile organisational structures, such as cross-disciplinary teams that can respond rapidly to emerging issues, or project-management techniques that enable adaptive planning, including short sprints, iterative design, and user feedback loops. In terms of procedures, this might mean streamlining the approval process for pilot projects or allowing teams more discretion to test innovations without requiring approval for every minor change. For example, the UNDP has moved away from a purely top-down approach to a more networked, responsive model that is flexible enough to accommodate its Accelerator Labs.
To continue addressing global challenges, multilateral organisations must innovate, or risk being rendered obsolete by more agile actors. Given their legitimacy, neutrality, and the trust they have built with governments and communities, these institutions are more likely to have their technological solutions adopted, scaled, and sustained. But first, they must adapt to the age of rapid innovation. — © Project Syndicate, 2025
Yana Gevorgyan is the director of the Group on Earth Observations Secretariat
