Search results
Introduction to Gender Bias in AI
Generative AI tools are already in use across the UN system for drafting, translation, situation analysis and planning, among other tasks. Left unexamined, the biases embedded in their outputs risk perpetuating the very inequalities our work seeks to address. The Global Digital Compact, adopted in 2024, placed AI at the centre of the international community's commitments to an inclusive digital future, and the first Global Dialogue on AI Governance was convened in Geneva in July 2026. While governance frameworks continue to take shape, these tools are already shaping our daily work.
This one-and-a-half-day workshop is part of the UNSSC programme on Advancing Future-Oriented Capabilities for UN Transformation. It is designed for UN personnel and Bonn-based ecosystem actors who work with, or are beginning to work with, Generative AI tools in humanitarian, development, peace and policy contexts, and who wish to develop a grounded understanding of how gender bias operates in those tools and what can be done about it in practice.
Use of these tools has outpaced the guidance and training that accompany them. In a 2026 UNSSC consultation with Women, Peace and Security practitioners, most respondents reported using Generative AI daily, and nearly all said they encounter gender-biased, misrepresentative or context-poor AI content sometimes or often. Practitioners also observed that the tools do not merely reproduce bias but can introduce it into text that did not contain it, and that output quality drops in languages and regions where training data is thin.
By the end of the workshop, participants will be able to:
- Explain in plain terms how Generative AI produces its outputs, and why fluent, confident answers can still be biased or wrong.
- Summarize the UN framework for the ethical and responsible use of AI at work and name at least three principles they must take into account when using Generative AI.
- Analyze specific example of gender bias in an AI-generated output: name the form of bias and explain its effect on accuracy or inclusion, including for women, minorities or speakers of under-represented languages.
- Apply a structured critical questioning process to AI outputs, keeping professional judgement and accountability with the human user rather than with the tool.
- Draft at least two concrete commitments for the responsible use of Generative AI in their own work or team.
The workshop follows a flipped classroom model. Before the workshop, participants are asked to complete a 90-minute online module and a short pre-participation questionnaire. The in-person sessions build directly on the module.
Day One, 9 December
How Generative AI works, and who it leaves out
How the technology produces its answers; where gender bias and related risks to accuracy, inclusion and language arise; and how to recognize them in real work products, from first principles to group work on realistic scenarios.
Day Two, 10 December
Human judgement in the loop: critical questioning for responsible use of Generative AI
Why judgement and accountability must stay with people; hands-on practice with the UNSSC Debiasing Coach, a tool designed to strengthen users' own critical questioning rather than correct text for them; and practical commitments for responsible use.
The workshop is designed for UN personnel and professionals from the wider Bonn ecosystem who use, or expect to use, Generative AI tools in their work and who wish to develop a grounded, practical understanding of how gender bias operates in those tools and what they can do about it. Please note that only Bonn-based UN personnel and professionals are eligible to register and attend the workshop.
The cost of this initiative are covered by Germany's Federal Ministry for Economic Cooperation and Development