09:00 - 09:15
|
Welcome and opening
|
09:15 - 09:25
| Opening keynote
|
09:25 -
10:05
| Session 1: The bigger picture – Child online protection in a changing world Children's lives online are changing faster than ever, and so are the risks they face as well as the measures designed to keep them safe. New technologies, new laws and shifting social norms are redrawing the child online protection landscape in every region of the world. This opening session takes stock of that landscape.
Voices from across the globe share how risks and harms are evolving, which initiatives are making a difference, and where technical standards can add the most value. The session also considers the outcomes from the
2026 Plenipotentiary Conference on Resolution 179 on Child Online Protection, to set the scene for the day's discussions. |
10:05 - 10:30
|
Session 2: Building children's rights into technology and standards Rights-based approaches balance protection with children's privacy, their access to information and services, and their participation in the digital world – a balance grounded in the UN Convention on the Rights of the Child and reaffirmed in the
Joint Statement on Artificial Intelligence and the Rights of the Child.
Building it into technical standards means considering children's rights at every stage of development, from age assurance and digital identity to AI. This session explores how rights by design can move from principle to practice, highlighting some considerations for standards bodies to take into account. |
10:30 - 10:45
| Coffee Break
|
10:45 - 11:00
| Keynote: "Mind the gaps" in child online protection standards
|
| 11:00 - 12:00 |
Session 3: Age assurance – Standards and the way forward Many online protection mechanisms begin with a simple question: is this user a child? Yet establishing a user's age reliably, while respecting their privacy, remains a central challenge for services, regulators and families alike. Answering it reliably, without collecting more personal data than necessary, is one of the defining challenges for online services, regulators and families. It is also an area where international standards have made real progress.
Recommendation ITU-T X.1901 "Age assurance systems – Framework", developed with ISO/IEC and technically equivalent to ISO/IEC 27566-1, gives the world a common framework for age assurance, with more work on the way. This session explores what these standards mean in practice, how they can work across services and borders, and how closer collaboration between standards bodies can take them further. |
| 12:00 - 13:00 |
Session 4: Privacy and protection – Encryption and parental controls As encrypted transport and technologies such as Encrypted Client Hello (ECH) become more widespread, they also change how child-protection measures that rely on network visibility can work, and how child sexual abuse material can be detected in end-to-end encrypted messaging.
Privacy and protection need not be a trade-off. This session explores technical approaches that deliver both, and how standards bodies can develop them together. Interoperable parental controls are one promising path:
Draft new Recommendation X.PARCEP "Interoperable PARental Control Enforcement Principles (PARCEP) for Child Online Protection", currently under development in ITU-T Study Group 17 alongside related work in the IETF, aims to let parental controls work seamlessly across devices and services which this session will examine in detail. |
13:00 - 14:30
| Lunch Break
|
14:30 - 15:30
|
Session 5: Child online protection in the age of AI – Model behaviour, agent identity and delegated authority Foundation models and agentic AI bring new challenges for child online protection. At the model level, the question is how a large language model or foundation model should behave when the user is a child – what it must refuse, escalate, safeguard against or never do – and how that behaviour can be benchmarked. This session explores a child-specific behavioural reference for benchmarking, building on ongoing work on security benchmarks for foundation models.
At the agent level, it examines identity and delegated authority when the human principal is a child: whether an AI agent inherits the child's age-assurance status, whether a child can confer authority on it, and under whose authority it acts, including questions of consent, authorization, accountability and revocation. |
15:30 - 16:30
|
Session 6: From standards to implementation – Building an effective COP ecosystem The true test of any standard is how well it works in practice. This session explores how standards, and the emerging requirements discussed earlier in the day, can be translated into safeguards that are deployable, interoperable and measurable. Taking an implementation-centred view, it identifies responsibilities, dependencies and barriers across the wider COP ecosystem, together with interoperability, conformance assessment, testing and measurable effectiveness.
|
16:30 - 16:45
| Coffee Break
|
16:45 - 17:15
| Session 7: What happens next – Priorities for standards This session consolidates the day’s findings and identifies priorities for future child online protection standardization.
|