As a global alliance, AUDRi brings a unique perspective that centres women, girls, and international human rights in order to advocate for safe, inclusive and equitable approaches to artificial intelligence. There are deep and serious harms caused by AI that disproportionately affect women and girls, from encoded bias to the use of AI tools to create deepfakes and other forms of abuse and exploitation. To counter this, women, girls and defenders of and advocates for gender equality must play a role in shaping the future of AI governance.
We have been working to shape the upcoming UN Global Dialogue on AI Governance. The Global Dialogue was created out of the Global Digital Compact and is a multi-stakeholder space for inclusive, evidence-based discussion on AI governance, enabling countries and organisations at different stages of digital transformation to align on norms, standards, and accountability mechanisms. The first dialogue will take place on the 6-7 July 2026 in Geneva. Contributing to this process allows us to influence global conversations and advocate for priorities that promote gender equality, safety, dignity and universal digital rights.
This process is particularly significant for AUDRi because it emerges directly from follow ups from the Global Digital Compact, which we used as a rallying point for recognition of the importance of gender considerations in digital governance. A key outcome of that earlier advocacy was the inclusion of a standalone action line on gender equality in the Global Digital Compact, an important milestone which came from our sustained civil society engagement through the Gender in Digital Coalition (GiDC). Gender must be treated as a structural pillar of digital governance rather than a cross-cutting afterthought, and, building on this foundation, AUDRi is now working to ensure that the Global Dialogue on AI Governance centres and mainstreams gender equality in its critical discussions and agreements.
Our work so far
- AUDRi submitted a formal written contribution to the Global Dialogue based on inputs from our global membership. These inputs reflect experiences across the Global Majority and highlight how AI systems are already shaping inequality, safety, and rights in everyday life. Through the inputs shared by our community we highlighted emerging issues such as concerns around digital colonialism, where data from the Global Majority is extracted and used in AI systems without consent, compensation, or meaningful local control. Environmental impacts and surveillance risks were also raised as issues that affect marginalised communities.
- In parallel, AUDRi is part of the Gender in Digital Coalition (GiDC), which also submitted a collective written contribution to the Global Dialogue process. This submission is fully aligned with AUDRi member inputs and reinforces many of the same priorities, particularly the need for gender-transformative AI governance.
- The official consultation was convened on 23 April and formed part of the preparatory process for the Global Dialogue on AI Governance and was hosted by UNFPA and the GiDC. There was attendance from the UN Secretariat. The consultation was designed to move Feminist AI from principle to practice, with a focus on dismantling structural barriers that prevent gender-responsive AI governance. It emphasised that gender equality in AI is not only about access, but about authorship and decision-making power over the systems that shape digital futures. AUDRi also contributed directly during the consultation, including in a dedicated session on safe and trustworthy AI, drawing on our member inputs and written submissions.
What are we asking for?
Through these interventions, we have articulated a shared set of priorities for transforming global AI governance from symbolic gender inclusion to structural accountability and gender justice.
1. Gender equality as a structural foundation of AI governance
A consistent message across all three interventions is that gender equality must be treated as a structural axis of AI governance, not an optional or thematic consideration. AI systems already shape access to employment, migration, healthcare, education, and public services, but governance frameworks often fail to reflect how these systems reproduce gendered inequalities.
The interventions collectively call for:
- Gender-responsive and intersectional governance frameworks
- Recognition of gender as a structural dimension of power in AI systems
- Operationalisation of gender equality through enforceable standards, not aspirational principles
2. Preventing harm through binding safeguards and safety-by-design approaches
We emphasise the need to move from reactive harm mitigation to preventive governance.
Key recommendations include:
- Mandatory pre- and post-deployment gender impact and risk assessments
- Safety-by-design approaches that embed protections at the earliest stages of AI development, and are rooted in consent.
- Stronger accountability mechanisms, including independent oversight and transparency requirements
- Consideration of moratoria on high-risk AI systems where safeguards are not yet in place
A particular focus is placed on AI-facilitated gender-based violence, including harassment, surveillance, deepfakes, and non-consensual content, which are framed as systemic risks rather than isolated incidents.
3. Structural inequality in global AI governance and data ecosystems
The interventions highlight deep fragmentation in global AI governance, where uneven regulatory capacity result in inconsistent protections across jurisdictions.
They also surface concerns about:
- Digital colonialism and extractive data practices
- Lack of consent, compensation, and data sovereignty for Global Majority communities
- Weak alignment between AI governance and data governance systems
A key recommendation is the creation of formal bridges between AI and data governance to ensure feminist principles of consent, collective rights, and sovereignty are embedded across both domains.
4. From token participation to meaningful power and co-design
A central cross-cutting theme is that participation must move beyond consultation toward shared decision-making and authorship.
Across all interventions, AUDRi and partners call for:
- Civil society speaking rights in plenaries and expert panels
- Guaranteed representation of feminist and Global Majority organisations in governance structures
- Structured review of outputs with civil society before publication
- Rebalancing participation from “presence” to “influence and authorship”
- Reordering dialogue formats so affected communities speak first, not last
This reflects a broader shift from inclusion as participation to inclusion as power.
5. AI grounded in human rights and lived experience
Across submissions and consultations, “safe and trustworthy AI” should consistently be defined not only in technical terms but in human rights terms.
This includes:
- Embedding human rights, transparency, and accountability throughout AI systems
- Developing survivor-centred safeguards for AI-enabled gender-based violence
- Ensuring human oversight in high-risk decision-making systems
- Addressing systemic bias and discrimination embedded in training data and deployment contexts
The consultation produced an outcome document that closely aligned with AUDRi and the GiDC’s written submission.
Looking ahead
As the Global Dialogue on AI Governance comes closer, AUDRi will continue working with its members and coalition partners to ensure that gender equality is not only acknowledged, but embedded across global AI governance systems. Including gender-responsive standards, strengthening accountability mechanisms, and ensuring that those most affected by AI systems are also those shaping their governance.
Together with the Gender in Digital Coalition will be hosting a pre-dialogue event in Geneva to engage with member states to help translate evidence and innovation emerging from Global South communities into concrete governance priorities, policy language, and implementation approaches that Member States and other actors can carry into ongoing AI governance negotiations and digital policy processes. We will be monitoring proceedings during the Dialogue and will be in attendance at the AI for Good Summit taking place the same week to monitor and contribute to conversations about AI principles and practice.
We are committed to advancing an AI governance ecosystem that is rights-based, gender-transformative, and grounded in the realities of the Global Majority. Follow for updates as we continue these important cross-cutting conversations!
