Session
Description:
The rapid advancement of artificial intelligence (AI) in security and defense is reshaping national and global security landscapes, yet governance frameworks remain fragmented. While AI ethics and governance in civilian contexts has gained momentum, reflected in the Global Dialogue on AI Governance and the Preliminary Report of the Independent International Scientific Panel on AI, the security domain lags behind, with limited multilateral, regional, and national engagement on responsible AI development and deployment in this space. Calls to move from principles into practice are growing, alongside recognition of the private sector's and investors' critical role in implementation and operationalization. Under the IGF 2026 sub-theme of Trust and Governance in the Age of Intelligence, this session has three objectives: (1) to assess what civilian AI governance can teach the security domain about structuring industry and investor engagement; (2) to surface emerging norms, policies, and legal frameworks for mitigating risk and ensuring accountability; and (3) to generate concrete recommendations for aligning AI development with global security norms while preventing misuse and enhancing international cooperation. Organized by the Roundtable for AI, Security, and Ethics (RAISE), a UNIDIR-led initiative convening AI developers, academics, civil society, international organizations, and policymakers across geopolitical divides, the session brings together OHCHR, the Kenyan Ministry of Defense, Microsoft, the Brazilian Presidency, and Privacy International. Outcomes will feed directly into the Fourth REAIM Summit (Kenya, April 2027) and the development of the Framework of Responsible Industry Behaviour for AI in the Military Domain, led by UNIDIR in partnership with OHCHR. This will ensure this workshop's insights concretely shape near-term multilateral and contributes to stability rather than exacerbating risks and inequalities.
The rapid advancement of artificial intelligence (AI) in security and defense is reshaping national and global security landscapes, yet governance frameworks remain fragmented. While AI ethics and governance in civilian contexts has gained momentum, reflected in the Global Dialogue on AI Governance and the Preliminary Report of the Independent International Scientific Panel on AI, the security domain lags behind, with limited multilateral, regional, and national engagement on responsible AI development and deployment in this space. Calls to move from principles into practice are growing, alongside recognition of the private sector's and investors' critical role in implementation and operationalization. Under the IGF 2026 sub-theme of Trust and Governance in the Age of Intelligence, this session has three objectives: (1) to assess what civilian AI governance can teach the security domain about structuring industry and investor engagement; (2) to surface emerging norms, policies, and legal frameworks for mitigating risk and ensuring accountability; and (3) to generate concrete recommendations for aligning AI development with global security norms while preventing misuse and enhancing international cooperation. Organized by the Roundtable for AI, Security, and Ethics (RAISE), a UNIDIR-led initiative convening AI developers, academics, civil society, international organizations, and policymakers across geopolitical divides, the session brings together OHCHR, the Kenyan Ministry of Defense, Microsoft, the Brazilian Presidency, and Privacy International. Outcomes will feed directly into the Fourth REAIM Summit (Kenya, April 2027) and the development of the Framework of Responsible Industry Behaviour for AI in the Military Domain, led by UNIDIR in partnership with OHCHR. This will ensure this workshop's insights concretely shape near-term multilateral and contributes to stability rather than exacerbating risks and inequalities.
Policy Question(s)
• What is the role and impact of the private sector and investors in governing AI in security, and how can this be leveraged to align development with global security norms, prevent misuse and enhance international cooperation?
• Where does responsibility sit across the value chain, from developers to State procurers to end-users? How can the Framework of Responsible Industry Behaviour clarify this?
• What would responsible AI procurement for security and defence look like in practice, in terms of due diligence and transparency?
• How can private-sector engagement strengthen multilateral governance and overcome fragmentation from company-specific or bilateral arrangements?
Expected Outcomes
• Heightened awareness of the private sector's and investors' role in operationalizing responsible AI governance in security and defence, and of the risks of leaving this engagement unstructured.
• Actionable policy recommendations for governments, international organizations, investors, and industry on what meaningful, accountable engagement should look like, including practical guidance on responsible procurement.
• Stronger multistakeholder collaboration through RAISE, deepening dialogue between policymakers, industry, investors, academia, and civil society on norms, standards, and accountability mechanisms for AI in security.
• A session output feeding directly into the Framework of Responsible Industry Behaviour for AI in the Military Domain (UNIDIR/OHCHR) and the Fourth REAIM Summit (Kenya, April 2027).
Format
Roundtable
Duration (minutes): 90
Format description: We are organizing this session as a roundtable to draw fully on both panelists' and attendees' expertise, and to structure active participation throughout. UNIDIR and OHCHR bring a strong track record of moderating interactive, dynamic panel discussions that engage speakers and audience alike; OHCHR's B-Tech Project will additionally support private sector buy-in. We have deliberately requested a 90-minute format to maximize discussion time among panelists and with participants more broadly. To ensure balanced online/offline participation, onsite and online moderators will coordinate closely, alternating questions between in-person and remote participants. The moderator will play an active role, steering discussion toward specific themes to surface the incentives, conditions, and success factors for meaningful and impactful engagement of the private sector and investors in the sensitive, yet highly consequential, domain of AI and international peace and security.
Duration (minutes): 90
Format description: We are organizing this session as a roundtable to draw fully on both panelists' and attendees' expertise, and to structure active participation throughout. UNIDIR and OHCHR bring a strong track record of moderating interactive, dynamic panel discussions that engage speakers and audience alike; OHCHR's B-Tech Project will additionally support private sector buy-in. We have deliberately requested a 90-minute format to maximize discussion time among panelists and with participants more broadly. To ensure balanced online/offline participation, onsite and online moderators will coordinate closely, alternating questions between in-person and remote participants. The moderator will play an active role, steering discussion toward specific themes to surface the incentives, conditions, and success factors for meaningful and impactful engagement of the private sector and investors in the sensitive, yet highly consequential, domain of AI and international peace and security.
Hybrid Format: The workshop will consist of a dynamic, moderated discussion among participants and speakers, with onsite and online moderators coordinating to alternate between in-person and remote participants.
To support this, organizers will circulate key guiding questions in advance as part of dissemination material, ensuring all participants arrive equally prepared for meaningful exchange.
Moderators will also use Pol.is, a web-based deliberation platform enabling direct engagement between in-room and online participants, a pioneering approach designed to close the gap between virtual and in-person contribution. Combined with polling tools such as Slido, participants will not only contribute to these collective intelligence activities but also react to and help consolidate emerging insights.
These insights, drawn equally from online and in-person participants, will feed directly into the session's outcomes and outputs, including the post-session report, and input toward the upcoming REAIM Summit and the Framework of Responsible Industry Behaviour for AI in the Military Domain.
Organizer 1: Yasmin Afina, UN Institute for Disarmament Research
Organizer 2: Isabel Ebert, United Nations Human Rights
Organizer 3: Jamal Hassan, Ministry of Defence, Kenya
Organizer 2: Isabel Ebert, United Nations Human Rights
Organizer 3: Jamal Hassan, Ministry of Defence, Kenya
Speaker 1: Jamal Hassan, Government, African Group
Speaker 2: João Brant, Government, Latin American and Caribbean Group (GRULAC)
Speaker 3: Ilia Siatitsa, Civil Society, Western European and Others Group (WEOG)
Speaker 4: Michael Karimian, Private Sector, Asia-Pacific Group
Speaker 5: Yasmin Afina, Intergovernmental Organization, Asia-Pacific Group
Speaker 6: Isabel Ebert, Intergovernmental Organization, Western European and Others Group (WEOG)
Speaker 2: João Brant, Government, Latin American and Caribbean Group (GRULAC)
Speaker 3: Ilia Siatitsa, Civil Society, Western European and Others Group (WEOG)
Speaker 4: Michael Karimian, Private Sector, Asia-Pacific Group
Speaker 5: Yasmin Afina, Intergovernmental Organization, Asia-Pacific Group
Speaker 6: Isabel Ebert, Intergovernmental Organization, Western European and Others Group (WEOG)
Moderator
Yasmin Afina, Intergovernmental Organization, Asia-Pacific Group
Online Moderator
Isabel Ebert, Intergovernmental Organization, Western European and Others Group (WEOG)
Rapporteur
Yasmin Afina, Intergovernmental Organization, Asia-Pacific Group
What will participants gain from attending this session?
