Beneficial AI for Children Coalition

Policy initiative

Recognizing a pivotal moment for future generations

People under the age of 18 represent 30% of our current world population and represent 100% of our future. During this critical stage of development, their brains are undergoing profound changes, where they are constantly learning new skills, accumulating knowledge and developing their perspectives on the world by interacting with their environment. This makes them especially receptive to opportunities but also deeply vulnerable to external influences — including digital technologies. Delivering AI approaches that support rather than undermine their cognitive, emotional, and social development is a legal and ethical obligation and an essential step toward our collective future.

AI fundamentally reshapes our world and offers significant new opportunities for expression, connection, and learning. For children, this could offer the potential to unlock more equitable access to education globally and across socioeconomic lines, and to provide personalized learning experiences tailored to each child's unique needs and abilities. AI can also help children globally realize their rights in specific context where they are threatened, such as their rights to education, to freedom of expression, of association and of peaceful assembly, to access information and culture, and to participate in relevant decision-making processes.

Harnessing AI's transformative power in education for the benefit of children demands deliberate action and requires tight collaboration with child development experts, educators and children. Unchecked digital media integration into children's lives can disrupt their development by affecting attention, altering delayed gratification and motivation, and influencing executive functioning and critical thinking. Generative AI's capacity to mimic human relationships can foster manipulative parasocial interactions in the context of children's immaturity, potentially leading to unhealthy attachments to AI agents and impaired social skills. As UNICEF notes, this balance underscores a dual challenge for children: those without access risk being left behind, while those with access remain vulnerable if AI is not developed with children’s best interests as a guiding principle.

This moment presents a rare and powerful opportunity to prioritize children's needs in the digital age. Understanding technology's power to empower and support young minds or to cause harm to their developing brains and well-being enables us to actively leverage research to ensure AI becomes a force for positive development.

We recognize that the trajectory of child development in the digital age is not yet determined, but rather a path we are collectively forging. The intentional and beneficial shaping of AI's interaction with children is paramount to ensuring they reach their full potential and well-being. Therefore, we are issuing this Call to Action to cultivate a unified understanding and commitment among diverse stakeholders to proactively address how to support and protect children’s healthy development in an age of AI.

Establishing an International Coalition to Safeguard Children's Development in the Age of AI

The Paris Peace Forum and everyone.ai are forming an international multistakeholder coalition that fosters inclusive and interdisciplinary dialogues between governments, institutions, tech companies, investors, researchers, educators, and families. Through diverse perspectives (spanning human sciences, human rights, technology, and ethics), we seek to support practical, evidence-based guidelines to safeguard and promote children's cognitive and socio-emotional well-being in the age of AI.

While this coalition specifically addresses the intersection of AI and child development, we acknowledge existing ongoing efforts related to online safety, data privacy, and digital literacy. This initiative, supported by the French Minister of International Partnerships and a number of governments, companies and civil society organizations, seeks to complement and build upon these efforts, providing a focused lens on the impacts of AI on developmental trajectories.

This coalition is guided by the following principles, intended to inform the design, implementation, and use of AI systems:

Child Safety: Prioritize and protect children's safety and well-being to foster optimal developmental skills, protect their privacy, and safeguard them from harmful and violent content and online risks in AI interactions including exploitation and abuse, while promoting trust and responsible AI practices that empower children.

AI that prioritizes the Best Interests of the Child: Uphold children's rights and foster AI solutions developed through a child-centered process that prioritizes their developmental needs. This commitment includes transparent practices and appropriate human oversight to align AI systems with children’s rights and well-being.

Ethical, Equitable, and Inclusive AI: Implement AI that strives to benefit all children regardless of background, ability, or needs, bridging the digital gap. AI systems must consistently mitigate bias throughout their life cycle to respect human rights and ensure equal accessibility.

Age-Appropriate AI by Design: Create and use AI solutions aligned with children's cognitive and socio-emotional stages accounting for differences in maturity and development across early childhood, adolescents, and older teens to promote their wellbeing.

Fair and Transparent AI: Communicate clearly the principles and processes behind AI decision-making for children, including data usage and training criteria, in ways children and guardians can understand.

Family-Inclusive Design: Provide caregivers with practical tools, and supportive resources so they can confidently safeguard children’s AI experiences. This also includes considering children’s perspectives in AI design, so that products align with their genuine needs and experiences.

Privacy for Children: Protect children's data privacy with robust safeguards, including informed choices, and that children are provided with transparent information about data processing, minimize personal data collection, de-identify data, and limit processing to what is necessary for product efficacy, improvement, ethical research and meet legal obligations the purposes pursued.

Objectives and Actions

To put our shared vision into action, we will collaborate in line with the rights of the child as enshrined in the Convention on the Rights of the Child and its three Optional Protocols to achieve the following objectives:

1. Establish, through continuous and structured dialogue including all stakeholders,shared guidelines that evaluate impact and mitigate risks of AI products used by children. This framework will be ongoing and evolve as our understanding improves. It will guide responsible AI development and deployment, prioritizing child safety and wellbeing while allowing for equitable opportunities for all actors.

2. Leverage scientific evidence for systematic benefit and risk assessment of AI products, and encourage long-term studies on the developmental, psychological, and societal impacts of AI on children.

3. Create a collaborative network of experts to provide ongoing guidance through dedicated committees and consultations, facilitating informed decision-making, fostering innovation, and ensuring AI solutions are aligned with children's developmental needs and rights.

4. Facilitate transdisciplinary international collaborations among all stakeholders while acknowledging each industry's unique strengths and challenges.

5. Promote AI education and literacy by developing comprehensive guidelines, pedagogical practices and educational programs that empower educators, parents, caregivers and children with the knowledge and tools to navigate AI safely and responsibly.

These objectives will naturally evolve over time to best reflect new insights and the changing landscape of AI.

We invite other organizations and individuals who share this commitment to join us in this critical endeavor by actively participating in collaborative efforts to shape a future where AI empowers and protects our children.

Our coalition

The participants in this coalition, co-led by the Paris Peace Forum and everyone.ai, to date:

- Tech companies: Open AI, Google, Anthropic, Hugging Face, PlugnPlay, Vivendi, evidence B

- Governments : France, Denmark, Norway, Chile, Mexico, Senegal, Bulgaria, Togo, Iceland

- NGOs, foundations and civil society: McGovern Foundation, Common Sense Media, 5Rights, Girl Effect, AI4good, Tralalère, Joan Ganz Cooney Center, Children and Screens, NSPCC, Edtech France, Respect Zone, AI4good, Fondation Enfance, We Protect, The Future Society, ICMEC, e-Enfance, InnovateEDU, Access to Knowledge for Development Center (The American University in Cairo)

UNICEF and UNESCO are also closely involved in this work.

The coalition also includes renowned experts such as Stuart Russell, Sonia Livingstone, Luc Julia, Gregory Renard,Mathilde Cerioli, David Harris, Laurence Devillers, Florence Gsell.