06 July 2026
IP26076 | Framing Expectations for the Fourth REAIM Summit
KEY TAKEAWAYS
• Kenya is set to host the fourth Responsible AI in the Military Domain (REAIM) summit in Nairobi from 7 to 8 April 2027.
• As host, Kenya will have a role to play in addressing some of the issues surrounding the REAIM summits, such as their value proposition relative to other multilateral initiatives and whether the summits should continue to produce outcome documents.
• Kenya’s success in addressing these issues at the upcoming summit will shape the platform’s future as well as the overall momentum among states to work together on military AI governance.
COMMENTARY
On 19 June 2026, Kenya announced that it will host the fourth Responsible AI in the Military Domain (REAIM) summit in Nairobi from 7 to 8 April 2027.
This announcement is significant for two reasons. First, it signals support among states to continue with the REAIM summits as a platform for multilateral dialogue on military AI governance. Second, it reinforces the REAIM summit’s position as an initiative led and sustained by middle powers.
However, there are some issues that need attention in the run-up to the next summit. The most important among these will be clarifying the REAIM summit’s value proposition relative to other multilateral initiatives on military AI governance, such as the resolutions passed at recent sessions of the UN General Assembly.
Related to this is whether the upcoming summit’s primary outcome will once again be a document endorsed by participating states. Nearly 60 states endorsed the 2023 inaugural summit’s “Call to Action”, while around 60 supported the “Blueprint for Action” produced at the second summit in 2024. In comparison, only 45 states endorsed the “Pathways to Action” document issued at the end of the third summit earlier this year.
Kenya’s success as the next host will depend on its ability to address these issues. It is also likely that the upcoming summit will be seen as a further test of the Global South’s ability to shape multilateral dialogue on AI governance, following India’s hosting of the civilian-focused AI Impact Summit in February 2026.
More than Muddling Through?
States that have lent their support to the REAIM summits since 2023 would probably have breathed a sigh of relief following Kenya’s announcement as the next host, along with confirmed dates for the event, as this cleared up the uncertainty regarding the initiative’s future following the third summit held in Spain in February 2026.
The uncertainty was caused by an inadvertent delay in holding the third summit, which disrupted the annual cadence as well as overall momentum among states to work together on military AI governance. Fewer states ultimately endorsed the “Pathways to Action” document compared to the “Blueprint for Action”, and both China and the United States failed to endorse a REAIM outcome document for the first time.
Nevertheless, the stage is now set for the REAIM summits to continue for at least another year, with middle powers at the helm of multilateral dialogue on military AI governance. The Netherlands and South Korea initiated and hosted the first two summits in 2023 and 2024, and the regional consultations held in the run-up to the second and third summits have fostered a model of inclusive multilateral dialogue.
Kenya’s previous experience as a co-host of the second summit held in Seoul in 2024 and convenor of the Africa regional consultations in 2024 and 2025 will help it to navigate the upcoming summit’s organisation. However, the issues that need attention in the run-up to the next summit have little to do with the logistics of organising a large event and will instead require Kenya to build consensus among the group of middle powers that have quietly come together in support of the REAIM summits over the past few years.
Clarifying REAIM’s Value Proposition
Front and centre among the various issues that Kenya will have to juggle will be the existential question of clarifying the REAIM summit’s value proposition relative to other multilateral initiatives on military AI governance. This has been a growing subject of discussion following the passage of Resolution 79/239 and Resolution 80/58 by the UN General Assembly in 2024 and 2025, respectively.
Both resolutions were led by the Netherlands and South Korea, the two states that initiated the REAIM summits, and have garnered overwhelming support. Resolution 79/239 led to a report published in June 2025 by the UN Secretary-General, which compiled the views of states and other stakeholders, while Resolution 80/58 supported the organisation of informal exchanges to discuss the Secretary-General’s report.
The informal exchanges were held recently in Geneva from 15 to 17 June 2026, with a broad agenda that covered topics ranging from foundational terminology to next steps for multilateral dialogue on military AI governance. Unlike the REAIM summit outcome documents, the UN Secretary-General’s report and the informal exchanges have not yet advanced specific areas of consensus among states.
There is thus value in retaining the REAIM summits as a platform for building consensus, particularly on technical and policy issues that would not be suitable for discussion at a platform like the UN General Assembly. Examples of such issues include the impact of AI on military decision-making, the application and enforcement of international law, and leveraging test, evaluation, validation, and verification practices to support regulation of military AI systems.
Going Beyond Another Outcome Document
In order to build consensus on these technical and policy issues, the upcoming summit should move away from primarily aiming to produce an outcome document. Three alternative ways forward include: creating working groups to develop guidelines on specific issues; providing a permanent platform to engage the private sector, which is often not at the table in multilateral dialogue platforms; and expanding the scope of regional consultations to deepen capacity building for a larger group of states.
The first suggestion draws on an idea originally intended to advance the United States’ Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy launched at the first REAIM summit in 2023. The Biden administration was only able to hold one plenary meeting in November 2024 among the endorsing states, during which three working groups focusing on assurance, accountability, and oversight were formed. However, the change in administration at the start of 2025 saw a pause in all efforts to advance the Political Declaration.
The second suggestion builds on a recommendation outlined in the final report of the Global Commission on REAIM published in September 2025 for creating a permanent dialogue mechanism for military AI governance, but focuses it on engagement with the private sector. Issues pertinent to the private sector that have received less attention at multilateral platforms so far, such as establishing norms of responsible behaviour and AI safety, can be the focus of the dialogue platform, if established.
Finally, expanding the scope of regional consultations to support capacity building at the regional level was highlighted during the 2025 meetings held in the run-up to the 2026 third summit. This will not only mean more substantive involvement of regional organisations at REAIM summits, but also a concerted partnership to ensure the wider adoption of best practices identified in outcome documents.
Looking Ahead
On paper, holding the next summit in April 2027 should give Kenya enough time to make substantive preparations, but a lot can change in the coming months. Conversations at the upcoming second session of the UN Group of Governmental Experts on lethal autonomous weapons systems in August and at the 81st session of the UN General Assembly’s First Committee (Disarmament and International Security) from October to November will definitely shape the next REAIM summit’s focus.
Manoj Harjani is Research Fellow and Coordinator of the Military Transformations Programme (MTP) at the S. Rajaratnam School of International Studies (RSIS).
KEY TAKEAWAYS
• Kenya is set to host the fourth Responsible AI in the Military Domain (REAIM) summit in Nairobi from 7 to 8 April 2027.
• As host, Kenya will have a role to play in addressing some of the issues surrounding the REAIM summits, such as their value proposition relative to other multilateral initiatives and whether the summits should continue to produce outcome documents.
• Kenya’s success in addressing these issues at the upcoming summit will shape the platform’s future as well as the overall momentum among states to work together on military AI governance.
COMMENTARY
On 19 June 2026, Kenya announced that it will host the fourth Responsible AI in the Military Domain (REAIM) summit in Nairobi from 7 to 8 April 2027.
This announcement is significant for two reasons. First, it signals support among states to continue with the REAIM summits as a platform for multilateral dialogue on military AI governance. Second, it reinforces the REAIM summit’s position as an initiative led and sustained by middle powers.
However, there are some issues that need attention in the run-up to the next summit. The most important among these will be clarifying the REAIM summit’s value proposition relative to other multilateral initiatives on military AI governance, such as the resolutions passed at recent sessions of the UN General Assembly.
Related to this is whether the upcoming summit’s primary outcome will once again be a document endorsed by participating states. Nearly 60 states endorsed the 2023 inaugural summit’s “Call to Action”, while around 60 supported the “Blueprint for Action” produced at the second summit in 2024. In comparison, only 45 states endorsed the “Pathways to Action” document issued at the end of the third summit earlier this year.
Kenya’s success as the next host will depend on its ability to address these issues. It is also likely that the upcoming summit will be seen as a further test of the Global South’s ability to shape multilateral dialogue on AI governance, following India’s hosting of the civilian-focused AI Impact Summit in February 2026.
More than Muddling Through?
States that have lent their support to the REAIM summits since 2023 would probably have breathed a sigh of relief following Kenya’s announcement as the next host, along with confirmed dates for the event, as this cleared up the uncertainty regarding the initiative’s future following the third summit held in Spain in February 2026.
The uncertainty was caused by an inadvertent delay in holding the third summit, which disrupted the annual cadence as well as overall momentum among states to work together on military AI governance. Fewer states ultimately endorsed the “Pathways to Action” document compared to the “Blueprint for Action”, and both China and the United States failed to endorse a REAIM outcome document for the first time.
Nevertheless, the stage is now set for the REAIM summits to continue for at least another year, with middle powers at the helm of multilateral dialogue on military AI governance. The Netherlands and South Korea initiated and hosted the first two summits in 2023 and 2024, and the regional consultations held in the run-up to the second and third summits have fostered a model of inclusive multilateral dialogue.
Kenya’s previous experience as a co-host of the second summit held in Seoul in 2024 and convenor of the Africa regional consultations in 2024 and 2025 will help it to navigate the upcoming summit’s organisation. However, the issues that need attention in the run-up to the next summit have little to do with the logistics of organising a large event and will instead require Kenya to build consensus among the group of middle powers that have quietly come together in support of the REAIM summits over the past few years.
Clarifying REAIM’s Value Proposition
Front and centre among the various issues that Kenya will have to juggle will be the existential question of clarifying the REAIM summit’s value proposition relative to other multilateral initiatives on military AI governance. This has been a growing subject of discussion following the passage of Resolution 79/239 and Resolution 80/58 by the UN General Assembly in 2024 and 2025, respectively.
Both resolutions were led by the Netherlands and South Korea, the two states that initiated the REAIM summits, and have garnered overwhelming support. Resolution 79/239 led to a report published in June 2025 by the UN Secretary-General, which compiled the views of states and other stakeholders, while Resolution 80/58 supported the organisation of informal exchanges to discuss the Secretary-General’s report.
The informal exchanges were held recently in Geneva from 15 to 17 June 2026, with a broad agenda that covered topics ranging from foundational terminology to next steps for multilateral dialogue on military AI governance. Unlike the REAIM summit outcome documents, the UN Secretary-General’s report and the informal exchanges have not yet advanced specific areas of consensus among states.
There is thus value in retaining the REAIM summits as a platform for building consensus, particularly on technical and policy issues that would not be suitable for discussion at a platform like the UN General Assembly. Examples of such issues include the impact of AI on military decision-making, the application and enforcement of international law, and leveraging test, evaluation, validation, and verification practices to support regulation of military AI systems.
Going Beyond Another Outcome Document
In order to build consensus on these technical and policy issues, the upcoming summit should move away from primarily aiming to produce an outcome document. Three alternative ways forward include: creating working groups to develop guidelines on specific issues; providing a permanent platform to engage the private sector, which is often not at the table in multilateral dialogue platforms; and expanding the scope of regional consultations to deepen capacity building for a larger group of states.
The first suggestion draws on an idea originally intended to advance the United States’ Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy launched at the first REAIM summit in 2023. The Biden administration was only able to hold one plenary meeting in November 2024 among the endorsing states, during which three working groups focusing on assurance, accountability, and oversight were formed. However, the change in administration at the start of 2025 saw a pause in all efforts to advance the Political Declaration.
The second suggestion builds on a recommendation outlined in the final report of the Global Commission on REAIM published in September 2025 for creating a permanent dialogue mechanism for military AI governance, but focuses it on engagement with the private sector. Issues pertinent to the private sector that have received less attention at multilateral platforms so far, such as establishing norms of responsible behaviour and AI safety, can be the focus of the dialogue platform, if established.
Finally, expanding the scope of regional consultations to support capacity building at the regional level was highlighted during the 2025 meetings held in the run-up to the 2026 third summit. This will not only mean more substantive involvement of regional organisations at REAIM summits, but also a concerted partnership to ensure the wider adoption of best practices identified in outcome documents.
Looking Ahead
On paper, holding the next summit in April 2027 should give Kenya enough time to make substantive preparations, but a lot can change in the coming months. Conversations at the upcoming second session of the UN Group of Governmental Experts on lethal autonomous weapons systems in August and at the 81st session of the UN General Assembly’s First Committee (Disarmament and International Security) from October to November will definitely shape the next REAIM summit’s focus.
Manoj Harjani is Research Fellow and Coordinator of the Military Transformations Programme (MTP) at the S. Rajaratnam School of International Studies (RSIS).
