IGF 2019 OF #33 Developing policy guidelines for AI and child rights

Subtheme
Description

This year marks the 30th anniversary of the adoption of the Convention on the Rights of the Child (CRC). As the global community reflects on progress made, one change to the world of 1989 will significantly impact the next 30 years for children’s rights: artificial intelligence (AI). Progress in AI systems, unprecedented amounts of data to train algorithms, and increased computing power are expected to profoundly impact life and work in the 21st century, raising both hopes and concerns for human development. A number of governments, businesses, civil society organizations and researchers are rightly concerned about the future of AI for societies. They recognize the current window of opportunity to lay down ethical and policy safeguards, and to practically develop the software, algorithms and data standards needed to maximize the benefits while limiting the risks of an AI future. Yet, in our initial findings, the impact of AI on children is largely missing from these concerns and efforts. It appears that for governments, the private sector and civil society, there is a policy vacuum for considered and practical guidance on AI and children (see https://medium.com/politics-ai/an-overview-of-national-ai-strategies-2a…). A key program for UNICEF will be the development of a policy guidance for governments, businesses, the non-profit sector and the Organization itself, which, if applied, will create environments that support the safe and beneficial use of AI systems for children’s development. UNICEF will lead the creation of the guidance in partnership with a range of external stakeholders, including the Berkman Klein Centre, the IEEE, and interested governments. Beyond creating guidelines, a key challenge is how to translate them into practice. Implementing guidelines poses difficult decisions, such as finding the balance between the right to privacy and the benefits of big data-fueled interventions that can protect children. The guidance will thus be piloted with policy makers in select countries for validation and learnings as broad principles are adapted to different country contexts. Such implementation will create case studies for other countries to learn from, as well as provide feedback that can inform subsequent versions of the guidance. Please see the attached concept note for more information on the overall project. As the IGF will take place during the project, the proposed Open Forum policy question is: What policy frameworks are needed to ensure that AI enables the realization of child rights and respects the key principles enshrined in the CRC? Between now and the IGF in November, UNICEF will co-host a AI and child rights workshop with the World Economic Forum in San Francisco in May, host a workshop in New York in June, and collaboratively develop the draft guidelines with governments. In the Open Forum UNICEF, along with the Berkman Klein Center, will present the workshop outcomes and draft guidelines – as a work in progress – and seek further input and collaboration with the audience. The session seeks diverse inputs and will also convene some of the few governments working on AI guidelines or principles that have a specific focus on youth. This critically important work will be strengthened by thematic and regional variations in the session, and contribute to putting child rights on the AI policy agenda. The work is directly linked to the Data Governance theme of the IGF, sharing emerging approaches to ensure the development of child-centric data governance frameworks at national, regional and international levels.

Organizers

UNICEF
Sandra Cortesi, Director of Youth and Media, Berkman Klein Center for Internet & Society, Harvard University

Speakers

Steven Vosloo, Policy Specialist - Digital Connectivity, Policy Lab, UNICEF Jasmina Byrne, Chief, Policy Lab, UNICEF Sandra Cortesi, Director of Youth and Media, Berkman Klein Center for Internet & Society, Harvard University Speakers from a diversity of governments working on AI policies, such as: Finland, Colombia (Armando Guio), United Arab Emirates (names to be confirmed in the coming months)

Online Moderator

Alexa Hasse

SDGs

GOAL 5: Gender Equality
GOAL 10: Reduced Inequalities
GOAL 16: Peace, Justice and Strong Institutions

1. Key Policy Questions and Expectations

QUESTIONS:

  • What role do human and child rights play for you when designing, developing, implementing, and evaluating AI systems?
  • Have you thought about child rights when developing and/or implementing AI strategies? If so, how? If not, why not?
  • How would policy guidance in this context be most useful (content, format, etc.)?

EXPECTATIONS:

  • Better understand the needs of government, business, and civil society when it comes to the design, development, implementation, and evaluation of AI systems upholding child rights.
  • Brainstorm ideas for how child rights can be promoted when creating AI policies. 
  • Gather inputs on how the policy guidance can be most useful to government and business (content, format, etc.).
2. Summary of Issues Discussed

There was broad consensus that:

  • There is a need to focus on the impact of AI on children, since children represent 1/3 of all internet users and are active users of AI-based systems.
  • Policy guidance is needed in this area to help policymakers and industry interpret ethical AI principles into practice for children.
  • Different regions and countries have specific needs and the way in which AI systems are provided and regulated for children need to be localized.
  • The ICT industry and big tech companies play a crucial role in how AI systems are used by and for children, and need to be part of these efforts.
3. Policy Recommendations or Suggestions for the Way Forward

UNICEF will lead the creation of an AI and Children Policy Guidance working with a range of external stakeholders, including the Berkman Klein Centre, the IEEE, 5Rights Foundation, the World Economic Forum, and interested governments. The draft policy guidance will be shared in June 2020. After that governments and companies will be invited to pilot the guidance and provide feedback.

6. Estimated Participation

An estimated 30 people attended the session in person. The following presented:

Sandra Cortesi, Director of Youth and Media, Berkman Klein Center for Internet & Society at Harvard University 
Steven Vosloo, Digital Policy Specialist, Office of Global Insight and Policy, UNICEF
Armando Guio, Fellow, Berkman Klein Center (and previous advisor to the Government of Colombia on their AI strategy)
Sabelo Mhlambi, Fellow, Berkman Klein Center
Karuna Nain, Global Safety Programs Manager, Facebook

The session was moderated by Jasmina Byrne, Chief, Policy Unit, Office of Global Insight and Policy, UNICEF