Policy Guidance on AI for Children: Pilot testing and case studies
Gathering real experiences from the field
4 minute read
In the UNICEF Policy Guidance on AI for Children we invited all governments and companies to pilot the guidance and openly share their findings about how it was used, and what worked and what did not. These real experiences from the field can inform and improve future versions of the guidance and inspire others to implement more child-centred AI. In this spirit, UNICEF will work with a diverse group of government and business “pilot partners”, identified with the support of our network and partners and based on their interest in bringing a child-centred lens into their AI policies and practices. Their findings will be documented as case studies to be published on our site in mid to late 2021.
The following initiatives were selected to illustrate a range of contexts in which AI systems and policies could be more child-centred. The aim for each pilot organization is to document how the policy guidance was used and, where possible, measure the resulting change in the form of a case study. The approaches taken, lessons learned and insights gathered will be one contribution to the global effort towards AI policies and systems that support children’s development. Please visit this page in the coming months for updates and further details on the project.
Policies and strategies
National AI Policy
Rwanda Information Society Authority
The Government of Rwanda is developing a national AI strategy which will be implemented by April 2021. The policy aims to equip government agencies and other stakeholders in the country to empower AI developers, citizens and users, and support the beneficial and ethical adoption of AI.
National Digital Strategy
The Government of Malta is formulating a National Digital Strategy for 2021-2024 that will provide a set of guiding principles and policy actions to advance the country’s digital economy in the areas of emerging technologies, particularly artificial intelligence. The new Digital Strategy aims for a more inclusive ecosystem to ensure that all segments of society are reached, including children.
Policy for child-centric AI for the cities of Lund, Malmö and Helsingborg
As Sweden is increasingly moving to embed smart technologies into governmental processes to enhance transparency, accessibility and productivity, the cities of Lund, Malmö and Helsingborg are working together to find appropriate ways to translate existing legislation and policies to benefit child rights across multiple cities, municipalities and regions in Sweden.
Responsible AI framework and assessment checklist
The global retail chain H&M will work to integrate greater awareness and focus on children’s rights in their Responsible AI Checklist, a tool that is utilized by their product teams and product owners, machine learning engineers, data specialists and others involved in developing or using AI capabilities within the company.
Understanding AI ethics and safety: A guide for the public sector
The Alan Turing Institute
The London based Turing Institute will utilize the draft policy guidance as a tool to engage with and inform the UK government, civil society and the public on the topic of AI and children. Their aim is to encourage civil servants who are interested in conducting AI projects to consider child-related AI ethics and safety as a first priority.
AutismVR is a virtual reality experience, from the Nigerian based start-up Imìsí 3D, that is intended to raise awareness and teach people how to interact with children on the autistic spectrum. The system utilizes natural language processing and is aimed at those who interact with pre-adolescent children on the autistic spectrum, including older children, caregivers and teachers.
Education Innovation Challenge Baseline Assessment Tool
Directorate of Science Technology and Innovation
The Government of Sierra Leone is developing an analytic Baseline Assessment Tool which aims to enable its users to view and assess different variables related to student literacy and numeracy. The tool is intended to benefit parents, policymakers, government decision-makers, implementing service providers and the general public.
Honda Research Institute Japan & Joint Research Centre of the European Commission
The Haru robot project aims to investigate approaches in AI and empathetic communication through the design of robot behaviours that promote child rights with a special focus on fairness and explainability. We aim to include children in all phases of the design process and to explore their needs, perceptions and behaviours in relation to social robots designed to interact with children in various settings.
Milli chatbot and Chillaa chatbot
Helsinki University Hospital
The psychiatry department of the Helsinki University Hospital has developed two chatbots, Milli and Chillaa, to offer young people information and support for anxiety and depression, as well as tips for positive thinking and guidance to help teenagers deal with social anxiety.
AI Nuora system predicting social exclusion
Fujitsu and Eksote
The purpose of the AI Nuora system, which utilizes predictive analytics, is to help identify young people at risk of social exclusion, thereby improving targeted intervention and reducing healthcare and other costs. The current target audience is teenagers; however, the system may be applied to any school-aged child with additional model training.
The CrimeDetector system automatically labels online incidents such as harassment and bullying on social media sites in Finland and Sweden, with the aid of a human in the loop. It is currently intended for use by the company's service team, including lawyers and psychologists.
Ministry of Social Development
The Government of Chile’s Ministry of Social Development has created a Childhood Early Alert System to prevent violations of children’s rights by identifying and prioritizing children and teenagers who demonstrate high risk factors. This preventive tool seeks to deploy early interventions to protect children’s rights.
Allegheny County Department of Human Services
The Department of Human Services in Allegheny County, Pennsylvania has developed HelloBaby, a system that employs a predictive risk model for families in the area. It calculates the risk of a newborn child being removed from their home and placed in foster care within the first three years of life, and informs County responses.
Piloting organizations commit to adhering to these terms.