Joint statement: The inaugural Global Dialogue on AI Governance must deliver for children

Download the letter

As UN Member States gather for the inaugural Global Dialogue on AI Governance, children's rights and voices must be central to their discussions.

Two years after the adoption of the Global Digital Compact, there is no shortage of principles, commitments, or guidance. Organizations, experts, and child-rights advocates from across the world have helped build a clear framework for protecting children in the age of AI.

The challenge is no longer knowing what to do. It is doing it.

States have obligations to protect children's rights in the digital world under international law and have reaffirmed their commitments. Five years ago, the UN Committee on the Rights of the Child's General comment No. 25 clarified how States' obligation to respect, protect, and fulfil children's rights apply in the digital world including in the context of AI.4 Since then, additional global instruments - including the UN Secretary-General's High-level Advisory Body on AI's Governing AI for Humanity, UN General Assembly resolution 78/187 on the Rights of the child in the digital environment, and the Joint Statement on Artificial Intelligence and the Rights of the Child - all point in the same direction: children's rights must be at the center of AI governance. What is needed now is political will.

Governments must turn commitments into action through effective regulation, strong safeguards, and meaningful accountability to ensure that children's rights are protected in practice.

Children cannot be an afterthought in AI governance

Children are among the earliest and most enthusiastic users of new technologies, yet they remain largely absent from national AI strategies and governance frameworks.

This gap matters. AI systems are already shaping children's lives: the information they see, the relationships they build, the education they receive, the services they access, and the decisions made about them.

As AI becomes more deeply embedded in society, children's rights, safety, wellbeing, and development must be built into systems from the start, not considered only after harm has occurred.

AI and Children's Rights

AI presents real opportunities for children. It can support learning, creativity, communication, and access to services, particularly for children who face barriers because of poverty, disability, geography, or discrimination. But these benefits cannot be assumed.

AI systems are being deployed at a speed and scale that far outpaces the safeguards needed to protect children. When systems are not designed, tested, and governed with children in mind, they can undermine their fundamental rights, including privacy, safety, health, education, access to information, and participation in public life.

Too often, commercial incentives take precedence over children's best interests and as a result, children face a range of serious and growing risks.

• Privacy violations: Children's personal data is routinely collected, shared, profiled, and used to train AI systems. Their images, voices, behaviours, and interactions are often treated as commercial assets rather than information deserving special protection. The concentration of children's data in systems over which they have little visibility or control raises serious concerns about privacy, accountability, and data sovereignty.

• Health and wellbeing: AI systems designed to maximize engagement can exploit children's developmental vulnerabilities. Features that encourage prolonged use or emotional dependency may affect identity formation, mental health, and social development. In particular, general-purpose AI chatbots generate acute risks when they are presented as companions or trusted confidants without adequate safeguards.

• Misinformation and disinformation: AI can generate and amplify false, misleading, and harmful content at unprecedented speed and scale. Recommendation systems often reward content that attracts attention rather than content that is accurate, increasing children's exposure to misinformation, disinformation, and extremist material.

• Bias and discrimination: AI systems trained on biased or incomplete data can reinforce existing inequalities and create new forms of discrimination. These harms often fall most heavily on children who already face barriers or exclusion, including girls, children with disabilities, LGBTQI+ children, children from minority communities, and children living in poverty or crisis-affected settings.

• Educational impact: The growing use of AI in education raises important questions about its long-term effects on learning, critical thinking, social development, and educational outcomes. The deployment of AI in schools should be guided by evidence of educational benefit, not assumptions about technological progress.

• Equity and access: Expanding access to AI is not enough. Without strong protections, AI can deepen existing inequalities and expose children to new forms of surveillance, exploitation, discrimination, and manipulation. At the same time, millions of children continue to face barriers to connectivity, digital literacy, language access, and inclusive education, limiting their ability to benefit from technological advances. New financial commitments are necessary to address this gap.

• Safety and dignity: AI tools are making it easier to create and distribute non-consensual intimate imagery, deepfake nude images of children, and AI-generated child sexual abuse material. The scale and speed of these technologies increase the risk of serious and lasting harm.

These risks affect everyone, but they are particularly acute for children, who are still developing and may lack the experience, knowledge, or power needed to understand, challenge, or seek redress for harmful automated systems.

Basic principles

We can no longer allow children to be collateral damage in the race to develop and deploy AI.

Violations of children's rights are not an acceptable price of innovation. In the global AI race, children have become test subjects for untested and unregulated AI - and the result is harm, not progress. Innovation does not have to depend on the commercial exploitation of children and the erosion of their privacy, agency, or critical thinking.

The risks posed by AI are not entirely new, and neither are the solutions. Like earlier technologies, AI is often deployed at scale before its safety and effectiveness have been adequately demonstrated. Too often, commercial incentives have been allowed to outrun accountability. This must change.

We urge governments to adopt and enforce a simple principle: AI systems should not be used by or affect children unless they have been designed to respect children's rights and independently shown to be safe.

Governments, regulators and legislators14 must fulfil their obligations under international law and their commitments under the Global Digital Compact by implementing safeguards throughout the development, deployment, and use of AI systems.

Parents and children should not be expected to bear the burden of managing risks embedded in AI systems. Responsibility should rest with the companies that build, deploy, and profit from these technologies. They must demonstrate that systems affecting children respect their rights and are safe before they are brought to market.

Recommendations

Children around the world are calling on governments to correct this imbalance and require age-appropriate design that protects their privacy, safety, and rights by default.

The inaugural Global Dialogue presents an opportunity to move from broad agreement to concrete action. To meet their obligations under the Convention on the Rights of the Child, States should adopt and enforce laws that:

• Requires precertification: Companies must demonstrate that systems affecting children respect their rights and are safe before they are brought to market. This includes requiring the business sector to embed safety and privacy by design and default and to undertake comprehensive child-rights impact assessments across the AI value chain and address identified risks before deployment. Where there is risk of harm, tech companies and legislators must apply the precautionary principle to protect children's rights, privacy, and safety.

• Prohibits the exploitation of children: Governments must ban manipulative design practices that exploit children's vulnerabilities (notably for commercial purposes), jeopardize their agency, and foster unhealthy dependency - including anthropomorphic chatbots and AI systems used to generate child sexual abuse material - unless and until businesses are able to demonstrate their respect for children's rights and safety. This also includes ensuring that children's images, voices, biometric data, educational data, and behavioral data are not used to train AI systems without explicit, free, and informed consent, strict necessity, child-rights safeguards, and effective oversight.

• Ensures accountability: Holding businesses accountable for infringements of children's rights facilitated by their products or services.

• Ends unlawful digital surveillance, particularly in commercial, educational, and care settings.

• Delivers regulatory oversight: Requiring businesses to maintain high standards of fairness, transparency, accountability, and oversight across the AI lifecycle. This includes requiring businesses to implement technical standards, industry codes, and terms of service that adhere to the highest standards of privacy, safety, and ethics, and ensuring compliance with them.

• Closes loopholes: Prevent weaker protections in some jurisdictions or across different categories of AI systems becoming a route for exploitation of children who hold equal rights wherever they are.

• Delivers transparency: Requiring businesses to provide age-appropriate child-friendly explanations to children, and parents or caregivers, of their terms of service.

• Offers redress: Requiring businesses - and states where appropriate - to provide children and parents or caregivers with prompt and effective remedies and to ensure child-friendly justice.

• Fosters children's participation: Support children's safe and meaningful participation in the governance, design, and development of AI systems that affect them, giving their views due consideration.

  Promotes AI understanding: Promoting proven safe and beneficial uses of AI, as well as critical technical and ethical AI literacy within a broader child rights approach to support children's understanding of AI possibilities and limitations as well as the commercial exploitation inherent to most systems. This also includes providing resources and support to teachers, parents, and caregivers.

The debate is not whether AI firms have a duty to protect children. It is whether governments are prepared to act on their obligations and the commitments they have already made.

Signatories

To view the list of endorsing organisations and individual signatories, download the PDF letter here.