This High-Level Conference took place online and person in Brussels on 12 December 2024. The conference was organised by Equinet, with the support of Unia, the Swedish Equality Ombudsman, the Norwegian Equality and Anti-discrimination Ombud and Luminate. It was aimed at Equality Bodies, European law and policy makers, civil society organisations, experts and activists working on the implications of AI systems for equality and non-discrimination.
Background
The EU Regulation on AI (“AI Act”), the world’s first legally binding, comprehensive legislation on AI, regulating the impact of these systems on health, safety and fundamental rights, is already in force. The AI Act could be a unique opportunity to strengthen equality protection, equipping Equality Bodies with essential information to better monitor, detect and address algorithmic discrimination and enhancing the effectiveness of protection through existing equality law with the strong enforcement and sanctions provided by the AI Act.
Equality Bodies have a rich record of experience, specific legal powers and mandates, covering both the private and public sectors, which makes them well-placed to ensure access to justice for those affected and develop impactful protection and prevention strategies. They have already started to effectively leverage their diverse powers to ensure AI systems protect and promote equality. For example, there is a growing body of cases, including insurance and banking algorithms, access to education, facial recognition technologies, dating applications, targeted job advertisements and the use of the infamous Syri system in the Dutch childcare allowances scandal. Some Equality Bodies are working with tech developers to ensure systems are equality compliant, while others are partnering with civil society to enhance the detection of algorithmic discrimination in the handling of discrimination complaints.
Equality bodies, therefore, have a key role to play in the future implementation of the EU AI Act (AIA). Indeed, there are several provisions in the AIA that foresee and enable the involvement of Equality Bodies and other national fundamental rights authorities, for example, through access to technical documentation and provision of technical assistance by market surveillance authorities (Article 77) and information and collaboration obligations for market surveillance authorities (Articles 73, 79 and 82). The powers that the AIA gives Equality Bodies and the concrete mechanisms for equality protection that it provides is a reflection of the well-documented far-reaching impact of AI-enabled technologies on equality and non-discrimination. The Act references “equality”, “discrimination” and “discriminatory” forty-eight times, including explicit mention of binding non-discrimination law and as central component of the definition of “systemic risk” in relation to General Purpose AI Models.
The present conference aims to explore how the Act’s implementation could contribute to safeguarding equality, examine the Act’s blind spots and identify potential strategies for addressing them through the involvement of civil society, Equality Bodies and other relevant public authorities and stakeholders. The conference will also present Equinet’s active involvement in the development of the future harmonized technical standards, which will give technical expression to the Act’s legal requirements and thereby enable its practical implementation by AI developers and deployers. Equinet has been participating in technical working groups of European standardisation organisations to ensure that the technical standards contribute to safeguarding equality and fundamental rights through embedding concrete rights-protecting measures and requirements.
Read more about Equinet’s work on Artificial Intelligence and Equality.
Objectives
The conference met the following objectives:
- Explore how the implementation of the AI Act can contribute to protecting and promoting equality.
- Highlight the critical role of Equality Bodies and civil society in the implementation of the Act and explore venues for strengthened cooperation between Equality Bodies and civil society.
- Examine and identify pathways for reinforcing the role of Equality Bodies in ensuring access to justice for those affected in the context of algorithmic discrimination.
- Advocate for collective and public approaches to redress in the context of algorithmic discrimination through the empowerment of Equality Bodies.
- Create bridges between Equality Bodies and key institutional players in digital regulation and policy and highlight the need for a coordinated and harmonized approach in national and European-level enforcement of the AI Act.
Agenda
09:30 – 09:40 Welcome
Patrick Charlier, Co-Director of Unia and Equinet Chair
09:40 – 09:55 Introduction: The unique added value of Equality Bodies in the algorithmic context
Together for equality, Stronger for all: the role of Equality Bodies in tackling algorithmic discrimination and protecting equality in the algorithmic context.
Samuel Engblom, Deputy Equality Ombudsman of Sweden
09:55 – 10:15 Keynote Speech
The implementation of the AI Act: opportunities for advancing Europe’s equality agenda.
Brando Benifei, MEP, Co-Chair of the European Parliament Group on the implementation and enforcement of the AI Act
10:15 – 11:00 Equality protection in the algorithmic context before and after the AI Act. What does the AI Act change?
- Explainer 1: Beginner’s guide to the AI Act – Brief overview of the architecture and underlying logic of the Act. Situating and clarifying the role of equality and fundamental rights protection in the Act through an interactive exercise.
Karolina Iwańska, Digital Civic Space Advisor, European Center for Not-for-profit law
- Explainer 2: Equality Bodies and the AI Act – Mapping and explaining specific provisions and mechanisms under the Act that have the potential to strengthen the institutional protection of equality through Equality Bodies.
Magdalena Maier, Legal and Policy trainee, Equinet Secretariat
Milla Vidina, Coordinator Equinet Working Group on Artificial Intelligence Senior Policy Officer, Equinet Secretariat
11:00 – 11:25 Coffee break
11:25 – 12:35 Using the AIA to fight algorithmic discrimination: how would this work in practice?
Practical examination of the different ways in which the Act’s technical requirements and legal obligations are likely to be applied by different actors in the AI value chain (e.g. tech businesses) and by the authorities involved in the chain’s regulatory monitoring under the AIA, namely Market Surveillance Authorities (Article 74) and Authorities Protecting Fundamental Rights (Article 77).
Presentation: How to use the Artificial Intelligence Act to investigate AI bias and discrimination: A guide for Equality Bodies (20 minutes)
Brent Mittelstadt, Professor of Data Ethics and Policy and Director of Research at the Oxford Internet Institute, University of Oxford
Discussion panel
Christiaan Duijst and Rik Helwegen, Dutch coordinating supervisor on AI and algorithms
Radboud van Delft, Netherlands Institute for Human Rights (designated Article 77(1) body under the AIA)
Patricia Shaw, Co-investigator in the “Equality by design, deliberation and oversight” Project, representing Equinet at JTC-21, Director of Women Leading in AI, CEO of Beyond Reach Consulting Limited
12:35 – 13:35 Lunch break
13:35 – 14:35 Protecting and promoting equality through the implementation of the AIA: how can the AIA live up to its potential?
Discussion and analysis of how applying the AIA into actual practice could further the protection of equality with a focus on identifying critical intervention opportunities and ensuring effective and accessible mechanisms for public oversight and accountability.
Kim Van Sparrentak, MEP, Member of the Committee on the Internal Market and Consumer Protection, European Parliament
Bjørn Erik Thon, Equality and Anti-discrimination Ombud of Norway
David Reichel, Project Manager of Social Research at the Research and Data Unit at the European Union Agency for Fundamental Rights (FRA)
Caterina Rodelli, EU Policy Analyst, Access Now
14:35 – 15:25 Using the AIA to strengthen access to justice against algorithmic discrimination: strengthening partnerships with civil society
Discussion centered on one overarching question: “How can we realize the principle ‘Nothing about us without us’ in relation to the regulation of AI technology?” The session explores how Equality Bodies further the realization of this principle through their different roles and activities and examines the need for collective and public approaches to redress in the context of algorithmic discrimination.
Berty Bannor, Bureau Clara Wichmann
Mher Hakobyan, Advocacy advisor on AI regulation, Amnesty International
Nele Roekens, Legal Adviser, UNIA and Chair of ENNHRI’s AI Working Group
15:50 – 16:35 Not a one’s law job: Leveraging synergies and building partnerships between equality law, data protection law and consumer law
Examination of the interfaces and interactions between three key legal frameworks implicated in the protection against AI-enabled discrimination. The discussion aims to identify practical pathways for strengthening inter-institutional collaboration between public authorities and facilitate knowledge exchange between the different legal areas.
Frederico Oliveira da Silva, Senior Legal Officer, BEUC – The European Consumer Organisation
Joanna Parkin, Legal Officer, Supervision and Enforcement Unit, European Data Protection Supervisor
Fabian Lütz, Equinet Legal Officer, Project “Equality-proofing AI systems: Equality by Design, Deliberation and Oversight”
16:35 – 16:45 Closing
Tena Šimonović Einwalter, Ombudswoman of Croatia
Resources
- Concept note and Agenda
- Karolina Iwańska – European Artificial Intelligence Act
- Brent Mittelstadt – How to use the Artificial Intelligence Act to investigate AI bias and discrimination: A guide for Equality Bodies
- Magdalena Maier, Milla Vidina – Equality Bodies and the AI Act
For further information, contact Milla Vidina (milla.vidina@equineteurope.org).