Gender & AI

Context

While artificial intelligence holds enormous potential to advance equality and inclusion, it has also amplified existing gendered harms in ways that are increasingly visible — from deepfake abuse and online harassment to algorithmic bias in hiring, credit scoring, and content moderation. These are not isolated technological accidents but the result of deeper structural inequities that are replicated and scaled through data, design choices, and deployment environments.

To move beyond reactive approaches, there is growing recognition of the need to look “upstream” — at how AI systems are conceived, trained, and governed from the outset. This means embedding gender-sensitive thinking throughout the AI lifecycle: from the composition of development teams to the selection of training datasets, the framing of use cases, and the ethical guardrails applied during model deployment. Proactive measures such as equity-by-design principles, diverse participation in AI research and governance, and robust gender impact assessments can help anticipate and prevent harms rather than merely respond to them.

By shifting attention from mitigation to prevention, and from representation to participation, we can begin to shape AI systems that not only avoid reproducing gender bias but actively promote fairness, inclusion, and empowerment. The challenge — and opportunity — lies in ensuring that gender considerations are not an afterthought in AI governance, but a foundational element of responsible innovation.

Description

This module explores the intersection of artificial intelligence and gender, recognising the gendered harms that AI systems can perpetuate—from deepfake abuse to algorithmic bias. Adopting a solutions-oriented perspective, with a focus on actions that parliaments can take, the session will highlight how ethics-by-design approaches, greater diversity in AI development, and equitable access to science, technology, engineering and mathematics (STEM) education can mitigate these risks and foster inclusion. Partners will also demonstrate how AI can be leveraged to detect and respond to technology-facilitated gender-based violence such as deepfakes, drawing on practical examples from across different contexts. The discussion will equip participants with strategies to advance equity, safety, and empowerment in the AI era.

Organization

• Scene-Setting: the current situation regarding existing gender biases and AI-enabled Technology-Facilitated Gender Based violence

• Roundtable discussion on “upstream solutions” and how to move away from reactive responses to gender harms and towards more proactive empowerment and engagement. 

• Showcase: Outline and Demonstration of the CSW GPT, with outline of the GPT’s origins and development, as well as encouragement of delegates utilizing the tool

• Interactive Exercise: Gender priorities