November 12, 2026 – The climate financing gap between what developing countries need for adaptation and what they receive remains wide. An estimated $305 billion a year by 2035 is needed — about 10 times the $28.2 billion to $31.7 billion in international public adaptation finance these countries received in 2024.

For least developed countries and small island developing states, the challenge is even greater. In addition to being disproportionately more vulnerable to the impacts of climate change, their adaptation finance needs amount to a median of about 3.6% of national GDP, twice the relative needs of upper-middle-income developing countries. Least developed countries and small island developing states also rely almost entirely on public sources such as the Green Climate Fund (GCF) to meet those needs.

To support these countries in this process, WRI’s Data Lab and the NDC Partnership Support Unit have built Project Ponte (which translates to “bridge” in Portuguese), an AI-assisted review tool that helps countries strengthen their access to climate finance.

While still in the prototype phase, Project Ponte makes the application process for GCF funding easier to navigate by helping countries strengthen their concept notes against the criteria GCF reviewers look for. It can also reduce the number of review rounds a proposal goes through, saving both time and money.

As noted in a recent policy brief published by the NDC Partnership Support Unit and RMI, the gap between national climate priorities and well-prepared, implementable and investment-ready projects remains one of the most persistent obstacles between climate ambition and delivery at scale.

Project Ponte aims to help least developed countries and small island developing states reduce the time and cost required to draft project concept notes, while also reducing their dependence on external consultants, often driven by their limited staff and climate risk data.

Project Ponte is an example of how AI can be part of the climate solution, in line with the Antalya Artificial Intelligence Pledge. Released by the COP31 presidency of Türkiye, the voluntary pledge calls for, among other things, stronger financial support and technical infrastructure for AI for climate action, particularly in developing countries, least developed countries and small island developing states.

Built Around What Countries Need

Data Lab’s Product Studio built Project Ponte based on thorough user research that highlighted countries’ and accredited entities’ needs in evaluating the readiness of their concept notes, rather than a tool that drafts it for them.

The tool was shaped by the experiences and insights from a working group of partner countries and institutions, including Fiji, Jamaica, Cambodia, Mauritius, Rwanda, Somalia, Belize, GCF, the Commonwealth Secretariat, the Global Green Growth Institute and the Climate Finance Access Network. Additional countries, including Benin, Madagascar, Haiti, Nepal and Mali contributed data to support the development of the tool.

The tool gives country users GCF-aligned feedback before concept notes are submitted and focuses on the climate rationale (or the climate change context) section, which users and experts both identified as one of the most common reasons proposals don’t advance. Ponte checks this section against the same hazard, impact and vulnerability criteria GCF reviewers use.

Our latest review criteria.
The tool checks climate rationale draft concept note text against five criteria to identify priority actions for author revision.

A recurring problem that users raised with using general AI tools, like ChatGPT or Claude, for this process is that they often received incorrect or invented data. Project Ponte avoids that risk by never rewriting text. Instead, a red, yellow or green signal for each theme tells users which text needs work first, flagging gaps and citing its sources. All feedback is grounded in current GCF guidance and the country’s own climate documents; the country is responsible for writing the content, keeping full ownership of it.

Project Ponte was tested with users from least developed countries and small island developing states, as well as with reviewers from the GCF. GCF colleagues shared that an expert is available to judge a project’s credibility early in the review process. Project Ponte aims to give that early signal to every country, not only those that can pay for expert advice.

Improving the Process for Countries and Their Partners

Beyond the interests of individual countries, the tool aligns with the goals of many actors across the GCF process: Each country has a National Designated Authority, the government office that links it to GCF and confirms that proposals match national priorities. These authorities can use Project Ponte to check their own drafts early without hiring consultants, or to review drafts from accredited entities, the national, regional or international organizations certified by GCF to develop and manage projects. For GCF, it could mean more consistent reviews. For partners and donors, the tool keeps decisions in countries’ hands, in line with GCF’s goal of giving national and regional institutions direct access to funding, while improving country users’ capacity to improve their own concept note drafts.

Get Early Access and Share Your Feedback

Feedback from users will continue to shape the tool’s development. The Project Ponte prototype is currently password-protected, with data loaded by hand for a small set of pilot countries that includes Antigua and Barbuda, Fiji, Cambodia and Cabo Verde. For those working on project development in these pilot countries, passwords can be requested via [email protected]. A public demo is also available on the Project Ponte website.

The prototype, which was funded by the government of the Netherlands, will be presented at an event during COP31 in Antalya, Türkiye, where the team will gather feedback and seek support for the next phase of the tool. Additional funding would keep the prototype running for pilot countries, add data for more least developed countries and small island developing states, and expand the tool to more concept note sections, mitigation projects, earlier readiness checks, team collaboration and exploring other funds, such as the Adaptation Fund and the Global Environment Facility.

Everyone is welcome to review the public demo and share feedback. Least developed countries and small island developing states are welcome to request to join the pilot. To compare approaches and lessons on building AI tools for proposal or compliance review, please get in touch.

Explore More:
data artificial intelligence (AI) COP31 National Climate Action