It Comes in Peace? AI’s promise and peril for Women, Peace and Security

Twenty-five years ago, the Women, Peace and Security (WPS) agenda called for women’s participation in peace processes and protection from violence.

Its anniversary is celebrated at a moment of rising global insecurity, simmering backlash against gender equality, and shrinking civic spaces. At the same time, fast-paced progress in Artificial Intelligence (AI) and cyber technologies promise to advance humanitarian response and transform our understanding of conflict.

I discussed this question at the 7th International Conference on Action with Women and Peace in Seoul, in a panel session with experts from the Korean Women’s Development Institute, Arab American University, UN Women, and the International Committee of the Red Cross. Our conversation uncovered how AI could, when implemented safely and transparently, support progress across all WPS pillars.

Participation: Expanding women’s voices in peace processes

AI can open new pathways to increase women’s inclusion in decisionmaking. Deliberative AI can ease participation by summarizing long text, reconciling alternative perspective and supporting iterative planning. Digital spaces can increase women’s participation in peace processes even in violent settings where physical access is constrained, such as in Yemen, where online platforms have enabled women to influence political outcomes without exposing themselves to physical danger.

Large language models can collect and systematize data, providing information that enables tracking progress on women's engagement. For example, the EMPOW project led by PRIO Research Director Louise Olsson leverages AI to collect and harmonize data on women’s mobilization across all stages of the peace process. This data can facilitate an understanding of what encourages or harms women’s participation.

Protection: Detecting and responding to threats

AI-driven analytics and biometric tools can help detect spikes in online misogyny, track hateful content, and identify patterns of conflict-related sexual violence. AI training initiatives, such as those run by Professor Demaidi in Yemen, could empower women to build solutions for sustaining livelihoods, addressing daily challenges in conflict settings and ultimately increasing their resilience. UN Women’s cyber resilience training across Bangladesh, Timor-Leste and many other countries equips women peacebuilders, journalists and human right activists with tools to monitor and respond to digital threats.

Prevention: Anticipating risks before they escalate

AI could strengthen conflict prevention by identifying early signs of gender-based violence before they escalate. AI tools can analyze vast streams of information to help peacebuilders anticipate risks and better allocate resources. For example, the Violence and Impacts Early Warning System (VIEWS) developed at PRIO and Uppsala University employs machine learning to predict the risk and intensity of armed conflict, and it is being expanded to predict the impacts of armed conflict on food security, income, health and other dimensions of development. As women and girls are disproportionately affected by the adverse consequences of conflict beyond the battlefield, these developments are highly relevant to the WPS agenda.

Relief and recovery: Enhancing humanitarian response

Lastly, deep learning and agentic AI can help map needs, monitor damage to infrastructure, and target humanitarian assistance more accurately. Drones and robotics may further improve the safe delivery of aid to women and girls in hard-to-reach areas.

The risks: Bias, exclusion and harm

Despite these opportunities, AI also brings significant risks. AI systems reflect the biases and exclusions embedded in their training data, and may worsen gender gaps. When women’s perspectives are missing from the teams that develop algorithms, the datasets used to train them, or the policy frameworks that aim to govern new technologies, AI-supported data and outputs can misrepresent women’s needs, or unintentionally reinforce gender biases and discrimination.

AI can be misused to harm and endanger women and girls. As Gaelle Demolis from UN Women noted during the panel, the majority of deepfakes globally target women, and more than 50% of women journalists have experienced some form of online abuse. In Southeast Asia, three in four women organization are subject to online harassment, and these figures are even higher for women in public office. In Sudan, Ukraine, Myanmar and Libya digital tools are being misused to harm, humiliate and silence women.

At the same time, humanitarian actors are grappling with the integration of AI into military and security infrastructures. AI is being embedded in decision-support tools and cyber operations, raising questions about transparency and oversight in life-and-death decisions. As Soyeon Shin of the International Committee of the Red Cross highlighted, uncritical reliance on AI for decisionmaking can lead to mistakes and misinformation, which may provoke panic, induce mass displacement, and increase gender-based violence. The effects on women and girls in conflict settings may be devastating.

AI and WPS need each other

AI could revive the WPS agenda but only if the two are integrated. AI requires inclusive governance to function effectively in conflict environments, while WPS needs digital and data-driven tools to remain relevant and impactful in a rapidly evolving security landscape. The WPS agenda can provide the framework needed to ensure that AI technologies reflect rather than erase women’s voices, and are aligned with humanitarian principles.

Integrating AI and WPS can help to ensure that AI is safer, fairer and more transparent, while making the WPS agenda more resilient, forward-looking and aware of the digital realities shaping conflict and security today. Research work at PRIO, including emerging efforts like the EMPOW project, is beginning to explore these dynamics, and will keep engaging with the intersection of AI and gender, peace and security.

What must happen next

This integration is possible, but far from guaranteed. Governments, multilateral organizations, donors and research institutes can start by strengthening women’s leadership and participation across both domains. This includes:

• investing in women and girls’ training on AI, data literacy and cybersecurity;
• supporting women researchers and practitioners working at the intersection of peacebuilding and emerging technologies;
• removing structural barriers that prevent women from entering and advancing in the AI field;
• enforcing ethical and gender-sensitive risk assessments to ensure accountability of AI systems deployed in fragile and conflict-affected settings;
• investing in data and knowledge infrastructure to ensure that data are properly contextualized;
• setting up oversight mechanisms so that AI tools and data used in early warning, humanitarian response or policy consultation do not inadvertently reproduce bias or reinforce exclusion;
• supporting initiatives that use technology to enhance inclusion, prevention and protection of women and girls, such as scaling models that allow women to participate safely in political processes through digital platforms, and supporting community-based innovation where women adapt AI tools to local needs.

The future of peace will be shaped in digital as much as physical spaces, and the WPS agenda must enter the digital space to remain relevant. Accountable governance, women’s leadership in peace processes, girls’ digital literacy, long-term investments in knowledge infrastructure, and routine ethical assessments will determine whether AI becomes an engine of empowerment or harm.

• Paola Vesco is a Senior Researcher at PRIO