Artificial intelligence is not a space beyond the law: children’s rights must guide the design, development and use of digital systems
A broad group of United Nations bodies, international organizations and entities working to protect children has called for a child rights-based approach to the design, development, deployment, use and governance of artificial intelligence. The statement stresses that children are not merely technology users or a vulnerable category in need of protection; they are independent rights holders whose best interests, views and evolving capacities must be considered in all AI-related decisions.
The call appears in the Joint Statement on Artificial Intelligence and the Rights of the Child, issued in November 2025 with the participation of the International Telecommunication Union, the United Nations Committee on the Rights of the Child, UNICEF, the International Labour Organization, the Inter-Parliamentary Union, UNESCO, the United Nations Interregional Crime and Justice Research Institute, the United Nations Office for Disarmament Affairs, the Office of the United Nations High Commissioner for Human Rights, the Special Representatives of the United Nations Secretary-General for Children and Armed Conflict and on Violence against Children, the United Nations Special Rapporteur on the sale, sexual exploitation and sexual abuse of children, and Safe Online.
The statement presents a comprehensive international framework for addressing the impact of AI on children. It brings together privacy and safety protections, the prevention of violence and exploitation, responses to discrimination and algorithmic bias, and guarantees relating to education, participation, access to information, justice and effective remedies.
This framework is particularly significant amid the rapid expansion of generative AI applications, recommendation systems, digital assistants and chatbots, biometric-recognition technologies, automated educational tools and systems used to make decisions about, evaluate or classify children.
Major opportunities accompanied by profound risks
The statement recognizes that rapid advances in information and communication technologies, including AI, are fundamentally changing the world in which children live and creating unprecedented opportunities to advance their rights.
When designed and used responsibly, AI can broaden learning opportunities, provide assistive tools for children with disabilities, improve access to information and services, support early risk detection, strengthen efforts to address climate change, and improve methods for preventing violence and exploitation.
These opportunities, however, are accompanied by major challenges that may affect children’s rights directly or indirectly. A child may personally interact with an AI system, but may also be affected by a decision taken by an algorithm without knowing that the system exists or understanding how it operates. Such decisions may relate to admission to an educational institution, assessment of academic performance, selection of the content shown to the child, classification based on personal data and online behaviour, or an assessment of the child’s need for social or health services.
The statement notes that most AI-supported tools and applications, as well as their underlying models and systems, were not originally designed with children, their well-being or their rights in mind. A lack of AI literacy is also not confined to children; it affects teachers, parents, caregivers, government officials, policy-makers and other relevant stakeholders.
The response therefore cannot be limited to telling children to use the Internet carefully. Responsibilities must be clearly distributed among the State, educational institutions, technology companies, system developers, parents, civil society and oversight bodies.
The Convention on the Rights of the Child as a binding reference in the digital environment
The statement affirms that the Convention on the Rights of the Child and its Optional Protocols constitute the principal legal framework for protecting children’s rights, regardless of the technology involved.
Under the Convention, a child means every human being below the age of 18. Children must be treated as a distinct group of rights holders, while taking account of their physical, social, emotional and cognitive development and their evolving capacities to exercise their rights with increasing independence.
The statement also builds on General Comment No. 25, issued by the Committee on the Rights of the Child in 2021, on children’s rights in relation to the digital environment. The General Comment confirms that children’s rights do not end at the boundary between the physical and digital worlds; they extend to all platforms, services and technologies affecting their lives.
Protecting children does not justify blanket censorship or depriving them of information and participation. Any restrictions on freedom of expression or access to information must be lawful, necessary and proportionate. They must reflect the child’s age and evolving capacities and must not turn protection into a pretext for excluding children from the digital sphere.
First: National governance based on children’s rights
The statement calls on States to adopt legislative, administrative, financial and policy measures that ensure AI governance respects, protects and promotes children’s rights.
This includes clearly identifying the responsible government entities, equipping them with the necessary powers and resources, and strengthening coordination among ministries, independent institutions, parliament, the judiciary, educational institutions, the private sector, civil society and children themselves.
Children’s rights should be integrated throughout the entire AI lifecycle: from data collection, model building and product design to testing, marketing, deployment, monitoring, updating and, where necessary, withdrawal.
The statement’s key recommendations include:
- Conducting periodic assessments of the risks AI systems pose to children’s rights.
- Publishing assessment findings and making them available to children in clear, age-appropriate language.
- Allocating sufficient budgets to protect children’s rights in the context of AI.
- Regularly reviewing policies and technical standards to keep pace with rapid developments.
- Ensuring access to child-friendly justice and effective remedies when violations occur.
- Engaging parliaments in legislation, oversight, budget approval and the accountability of public and private actors.
No automated system should be permitted to make consequential decisions about children without meaningful human oversight, the possibility of challenging the decision and access to an understandable explanation of the grounds on which it was made.
Second: Responsibility, transparency and accountability
The statement places responsibility on States to protect children from violations committed by third parties, including companies that design, develop, deploy or operate AI systems.
States are called upon to establish legal frameworks defining the civil, administrative and criminal responsibility of natural and legal persons, proportionate to the nature and seriousness of the act or omission and the resulting harm.
The technical complexity of AI must not become a means of evading responsibility. If an algorithmic system discriminates against a child, exposes the child’s data, subjects the child to harmful content or facilitates exploitation, those responsible must be identified, investigated and held accountable, and the child must have access to reparation.
Companies should also provide clear information about how their systems work, the data they collect, how those data are used and the criteria used to rank content or make decisions.
The statement stresses the need for accessible and safe mechanisms through which children, parents and caregivers can report problems and receive a prompt and meaningful response. Companies should conduct regular audits and child rights impact assessments, including tests of whether their systems can prevent the production of harmful, discriminatory or age-inappropriate material.
Third: Protecting children from violence and exploitation
The statement devotes substantial attention to safety and calls for the prevention of all forms of violence and exploitation committed through, or with the support of, AI systems.
Potential risks include:
- Psychological and sexual violence, including gender-based violence.
- Cyberbullying, harassment and sexual grooming.
- The creation or modification of child sexual abuse or exploitation material using AI.
- Deepfakes and other deceptive synthetic media.
- Content inciting hatred, violence or self-harm.
- Content promoting eating disorders, drug use or gambling.
- Disinformation and misinformation targeting children.
- Economic exploitation, child labour, trafficking and forced begging.
- Recruitment and use of children in armed conflict.
The statement calls for the explicit criminalization, investigation, prosecution and appropriate punishment of online sexual exploitation of children facilitated by AI, including AI-generated or AI-modified abuse material and grooming for the purpose of committing a sexual offence.
Digital platforms, gaming companies, streaming services, educational technology providers and social media companies should adopt appropriate age-assurance mechanisms when necessary and proportionate, while respecting privacy and avoiding the collection of unnecessary additional data.
The statement also warns about AI systems that imitate human interaction. Such systems should include safeguards preventing unhealthy emotional attachment by children. A child may perceive a chatbot as a friend, adviser or real person even though its responses are generated by an automated model that possesses neither human understanding nor moral responsibility.
Fourth: Privacy and data protection
The statement emphasizes that children’s privacy is a fundamental right, not merely a technical option or a clause buried in a lengthy terms-of-service document.
Legislation and policies must protect children’s data by design, provide independent oversight and ensure access to remedies. Privacy policies should be presented in language children and parents can understand, explaining what data are collected, why they are used, with whom they may be shared and how long they will be retained.
AI systems must comply with the principle of data minimization by collecting only what is necessary and avoiding the retention of information that the service does not require. Personal and biometric data must be protected through encryption, secure storage and regular audits.
A particularly serious concern is the use of children’s images, voices, online activity, conversations and schoolwork to train AI models without their knowledge or informed consent.
The statement also calls for protection from commercial models that turn children’s data, attention and digital activity into monetizable resources, creating detailed profiles of their preferences, habits and vulnerabilities for advertising or content targeting.
Fifth: The best interests of the child as a primary consideration
The best interests of the child must be a primary consideration in all AI-related decisions and actions, whether in the public or private sphere.
The concept must not be reduced to technical safety. It includes the child’s physical, mental, psychological, moral and social development, as well as the effective enjoyment of all rights recognized under the Convention.
Tools intended for children must be age-appropriate, communicate in language understandable to children and parents, and be developed through multidisciplinary cooperation involving specialists in technology, education, psychology, social work, law and children’s rights.
A single model cannot be presumed suitable for all children. What is appropriate for a six-year-old may not be suitable for a 16-year-old. A system accessible to a child without a disability may remain unusable for a child with a visual, hearing, intellectual or other disability.
Sixth: Non-discrimination and inclusion
AI can reproduce existing discrimination and may even expand it while giving it the appearance of technical neutrality.
If a system is trained on biased, incomplete or unrepresentative data, it may discriminate on the basis of gender, disability, origin, race, language, religion, socioeconomic status or place of residence.
The statement therefore calls for human rights due diligence to prevent and mitigate algorithmic bias and for systems and services to be made available in multiple languages and adapted to diverse cultural contexts.
The benefits of AI should be accessible to all children, including children with disabilities, girls, children living in rural or remote areas, and those in vulnerable or disadvantaged situations.
Addressing the digital divide requires more than providing devices and Internet access. It also demands affordable and reliable connectivity, digital skills, appropriate content and effective protection from harm.
Seventh: Children’s right to participate
Policies affecting children must not be developed without listening to them.
The statement calls for the safe and meaningful participation of diverse groups of children at every stage of AI policy development and in the design, development, deployment and review of AI models and products.
Participation must not be symbolic. Children should receive the information and support necessary to express their views safely, and those views should be given due weight in accordance with their age and evolving capacities.
Children are often best placed to describe their actual experiences with platforms, games, educational robots and recommendation systems. Their participation can reveal risks that conventional technical testing may overlook.
Eighth: Education and AI literacy
The statement recognizes that banning AI tools or ignoring their existence does not constitute an adequate education policy. Responsible integration is needed where appropriate and supported by evidence, alongside the development of the technical and critical skills children need to understand and use AI.
AI literacy should enable children to:
- Understand the basic principles through which algorithmic systems operate.
- Distinguish between human-created and AI-generated content.
- Verify information rather than automatically treating AI output as accurate.
- Protect their personal data.
- Identify misleading content and deepfakes.
- Understand how algorithms influence their choices and the information shown to them.
- Know their rights and how to report harm, challenge decisions and obtain assistance.
These competencies should be integrated into school curricula and non-formal education programmes. Particular attention should be paid to marginalized children to prevent a new divide between those who possess AI skills and those who are denied them.
Teachers, social workers, psychologists, police officers, prosecutors, judges and all professionals working with or for children should receive appropriate training. Government officials, parliamentarians, AI designers and developers, and corporate leaders must also be trained.
Ninth: AI and the environment
The statement connects AI governance with the child’s right to a clean, healthy and sustainable environment.
AI can improve climate modelling, energy efficiency, resource management and disaster preparedness. At the same time, some AI systems consume substantial amounts of electricity and water, while their infrastructure depends on raw-material extraction, manufacturing and data centres with significant environmental effects.
The carbon footprint, energy and resource consumption, and impact of raw-material extraction must therefore be assessed and mitigated. AI development must not undermine the rights of children today or those of future generations.
Tenth: Protecting children in armed conflict
The statement addresses AI use in the military domain and calls for safeguards consistent with international humanitarian law and international human rights law.
Systems used in decisions relating to hostilities or the use of force must remain subject to strict human oversight and control, with human judgment at the centre of decision-making.
AI technologies must not be weaponized or misused in ways that disproportionately affect children, particularly in conflict zones and in connection with autonomous systems or automated targeting tools.
Priorities for Lebanon
In light of the statement’s principles, Lebanon needs a national framework to protect children’s rights in the context of AI and the wider digital environment. Children, parents, public institutions, civil society, experts, educational institutions, and technology and telecommunications companies should all participate in its development.
Priority measures include:
- Integrating children’s rights into any national AI strategy, rather than limiting the discussion to innovation, investment and cybersecurity.
- Reviewing legislation on privacy, data protection, cybercrime and juvenile protection to ensure that it responds effectively to emerging risks.
- Requiring public and private actors to conduct child rights impact assessments before introducing AI systems in education, health, social services or justice.
- Adopting clear rules for AI use in schools, including the protection of students’ data, academic integrity, teacher training and a prohibition on exclusive reliance on automated decisions.
- Establishing confidential, child-friendly reporting mechanisms and providing psychological, social and legal assistance to victims.
- Strengthening the capacity of the judiciary and competent authorities to investigate AI-facilitated crimes, particularly sexual exploitation, deepfakes, extortion and privacy violations.
- Ensuring that children with disabilities and children in disadvantaged areas have access to safe and appropriate technology, in accordance with universal-design principles.
- Systematically involving children in policy development and in the assessment of services and platforms intended for them.
- Launching national digital and AI literacy programmes for children, parents, teachers and social workers.
- Strengthening independent oversight and public accountability, rather than leaving the protection of children solely to companies’ terms of service and internal policies.
Towards technology that serves children’s rights
The central question is no longer whether children will use AI. It has already become part of their educational, social and cultural environment. The real questions are under what conditions they will use it, who will establish those conditions and who will be held responsible when harm occurs.
The Joint Statement on Artificial Intelligence and the Rights of the Child offers a clear roadmap: technology must serve people, and its success must be assessed in light of its impact on those most exposed to risk, foremost among them children.
Technical safeguards alone are insufficient. Effective legislation, independent institutions, human oversight, corporate accountability, critical education, genuine child participation, access to justice and effective remedies are all required.
Protecting children in the age of AI does not mean isolating them from technology. It means enabling them to benefit from it safely, with dignity and on an equal basis, while respecting their privacy, voice, best interests and all rights guaranteed under international law.
However advanced AI systems may become, they cannot replace human responsibility. States, companies, institutions and communities remain responsible for ensuring that technological progress does not create new forms of discrimination, exploitation or exclusion, but instead becomes a means of advancing the rights and protecting the future of every child.
هذه المقالة متاحة أيضًا بـ:
