Brussels

Advancing more democratic AI governance with policymakers

General Goal & Structure

The overall purpose of the Brussels Use Case (UC1) is to identify central questions underlying AI governance, identify emerging priorities and understand how European policymakers can and should regulate the disruptive nature of general-purpose AI democratically.

In recent years, the EU has been active in positioning itself as a global regulatory leader on AI beyond enacting new policies and regulation. As national AI legislation and guidelines emerge beyond Europe, the EU finds itself at a pivotal moment in determining its technological stance on the global stage. How to uphold democratic values amidst intensifying AI progress and geopolitical competition? Emerging technologies such as AI not only raise questions in terms of their regulation, but also how they affect governance itself.

Primary Objectives

The Brussels-based Use Case will be centred around the concerns, discursive norms and values of policymakers, who are responsible for finding appropriate compromises between differing positions regarding the socially optimal outcomes and the need to keep regulation generalisable.

It seeks to make the policymakers and AI governance experts aware of the KT4D framing of issues of advanced knowledge technologies and democracy, and of soliciting from these individuals and groups their input to KT4D’s research processes.

The overall objectives of UC 1 are as follows:

How policymakers can regulate the disruptive nature of AI & big data democratically
How to design governance according to current speed of AI advances
Position EU as the global regulatory leader in AI
Critically evaluate how democratic values of integrity, transparency and privacy are accounted for in policy development
Capture negative impact of these technologies on fundamental rights and democracy, ensuring public trust in AI

Methodology

The four use cases follow a common methodology, based upon three interaction points between the project and their target stakeholder groups.

The first interaction was based upon the model of Participatory Design.
The second interaction will be constituted as a one-day Digital Democracy Lab.
The final interaction will be between the KT4D team and target groups, where pre-final versions of KERs will be available for interrogation and discussion.

The participant groups are built up out of the host partners’ current projects and networks, and have been assigned to localities to capitalise on partner networks and local community assets, such as the presence of policymakers in Brussels.

Interaction with other Use Cases

The Brussels Use Case interacts with the other Use Cases in the following ways:

Similar Combined Use Case Methodology
Coordination of the meeting events ensuring coordination between the contributing WPs, events management, including Use Case consultations and co-creation meetings

However, it maintains a connected relationship with Dublin (UC 4) in the following ways:

Co-creation of an agenda, discussion guide, and interaction design food the first interactions of the groups
Interaction design for final Valourisation Meetings

Target Audience

Policymakers

CSO’s

Citizens

Deliverables / Outcomes

Social risk toolkit (KER 1)

The Social Risk Toolkit provides a comprehensive overview and in-depth research insights covering the key socio- cultural issues associated with the intersection of AI, Big Data and Democracy. The toolkit consists of 8 individual modules covering different cultural or social aspects of AI and big data in democratic society.

Governance framework, policy roadmap and recommendations (KER 2)

The framework for democratic AI governance critically builds on existing ethical AI guidelines to support democratic deployment of AI in society. The framework and associated policy recommendations provide guidance on assessing the risks and opportunities AI poses to democratic participation.

Digital Democracy Lab Handbook & Demonstrator (KER 5 & KER 6)

A one-day experience of working within a transparently constructed, open and explainable platform for civic interaction. During the workshop, each of the core components will be explained and engaged with critically by the users in the lab in order to achieve civic interaction and actionable insights. Participants will be guided through a series of hands-on exercises working with the demonstrator platform and actively exploring the emerging ethical and technical issues. Each exercise will be accompanied by an opportunity for reflection on the affordances and constraints of the platform.

Updates & Recent News

Use Case 1 - Meeting 1

26 October 2023

Press Club Brussels Europe

While the shared goal of all the Participatory Design Sessions is to understand current and desired future user practices so as to align design to their requirements, each Use Case has a peculiar structure and a set of goals designed to best suit the participants’ needs, interests, and skills.

The first interaction was a roundtable instead of an intense workshop because the aim was mostly to tease out discussion, inputs and insights on AI regulation and policy instead of co-creating solutions at this point.

The overall purpose of the first interaction was to:

Explore and validate the risks and opportunities AI poses to democracy.
Identify blind spots and gaps in the existing EU governance approach to AI with respect to democracy.
Explore and gather existing positive examples of possible governance responses to AI systems, including general-purpose systems.
Crystallise questions for a future Delphi study on AI governance.
Analyse democratic governance of AI within European context.
Address challenges related to power centralisation, lack of transparency, and the dynamic nature of technological changes.

Use Case 1 - Meeting 2

15 October 2024

Press Club Brussels Europe

The goal of the workshop was to map pathways for responsibly utilising AI as part of deliberative processes and identify gaps in current AI governance approaches regarding democratic processes. The Beta version of Digital Democracy Demonstrator was presented to elicit feedback on initial guidelines for democracy-in-the-loop AI applications and democratic stewardship of the technology. The Digital Democracy Demonstrator only acted as a boundary object guiding the discussion.

The participants gained insights from the latest policy Delphi study findings on AI governance approaches in democratic contexts, and engaged with the Digital Democracy Demonstrator as a design research prototype, exploring practical applications of AI in democratic innovation.

The meeting was composed of three sessions:

Setting the Scene - AI and Democratic Innovations.
Engaging with the Digital Democracy Demonstrator.
Deepening the Discussion - AI Governance in Democratic Contexts.

Use Case 1 - Meeting 3

7 November 2025

Press Club Brussels Europe

The third meeting of UC1, titled ‘Expert Roundtable on Democratic AI Policy’, was carried out in Brussels, and it specifically aimed at developing the governance framework and a roadmap for regulating the use of AI and big data in support of democracy (KER2). It sought to clarify the different interpretations of democratic AI governance and the technical imaginaries underlying the notion. The use case aimed to engage experts and policymakers to co-create the key elements of the framework and validate analysis of the intersection between AI and democracy from the perspective of future-proof technology governance. The methods and data for crafting KER2 included:

A literature review of the relevant EU digital policy developments, technology governance and the intersection of AI and democracy.
Notes and outcomes from the three UC1 events in Brussels.
The results of the Delphi survey, conducted from spring to summer of 2024.

The roundtable resulted in valuable feedback and concrete draft policy roadmaps to inform drafting of the framework for democratic AI governance (KER2):

Refine the Governance Framework: The framework should explicitly show how AI interacts with the pillars of democracy and the AI lifecycle. Furthermore, policymakers need a clearer classification of required actions separated into short-term and long-term goals.
Strengthen Digital Sovereignty: Regional and national digital strategies must be developed to protect European interests and counter the disproportionate influence of foreign technology.
Focus on Enforcement and Capacity Building: Instead of solely drafting new rules, emphasis should be placed on the rigorous national-level enforcement of existing regulations, such as the AI Act, DSA, and GDPR. To achieve this, public institutions need greater technical capacity and resources, and SMEs require additional support.
Boost Citizen Engagement and Literacy: Improving digital and AI literacy across all generations is a primary requirement. Policy processes should also integrate genuine citizen participation mechanisms that give users stronger control over their data and services, rather than relying on weak 'opt-out' options.
Tackle Digital Injustice: AI policy must address the global environmental and ethical impacts of the technology.

The final framework for democratic AI governance (D5.1) also includes additional policy recommendations to implement the framework. To this end, the insights from the Delphi survey about the feasibility, probability and desirability of different AI policies have been used to support the evaluation of the measures proposed by the use case participants, and to formulate actionable recommendations.