Embracing the future: The EU AI Act

With the recent adoption of the EU AI Act, the European Union has taken a monumental step towards shaping the future of our continent. This European groundbreaking regulation establishes a solid foundation for the responsible and trustworthy adoption of AI systems. For local authorities, it represents a powerful tool to spur innovation, while keeping their territories and citizens at the centre of their policy action.

Are policy makers ready to take up the opportunity?

Allegedly, the answer to this question could be a simple “no”. In many contexts, the perception of AI can be rather negative: it’s “threatening”, “intrusive”, “unreliable”…the feeling of mistrust can be experienced by both citizens and public servants. This is the first challenge for local governments: opening up to the potential of an unparalleled and life-changing technology requires overcoming negative perceptions first, including among civil servants within their own organisations.

In policy making, tangible risks are also to be taken into account: privacy violation risks through potential data misuse; surveillance risks due to over-monitoring that may infringe on freedoms; AI could also entail decision-making biases leading to discriminatory practices in local governments, thus reinforcing existing inequalities. Finally, public bodies are concerned by the risk of supporting AI investments resulting in market failures bringing negative net social benefits and leading to social inequalities.

Understanding the EU AI Act

A close look at the AI Act shows that there are responses to the above potential concerns.

First, the Act provides clear guidelines prioritising transparency, accountability, and human-centric AI applications. All AI technologies must be developed and deployed in a manner that respects democratic principles and protects the rights of citizens.

Second, the Act classifies AI systems into four categories, under a risk-based approach: unacceptable risk, high risk, limited risk, and minimal risk. For instance, for high-risk environments, such as critical infrastructures, healthcare, and law enforcement, AI systems will be subject to strict conformity assessments, ensuring that they meet the highest standards of safety and ethical considerations.

Third, the Act provides a harmonised regulatory framework. Standardised rules create a fair landscape, encouraging public authorities to invest in AI technologies without the fear of regulatory fragmentation. This creates a conducive environment for the development of cutting-edge solutions that address specific local challenges and improve public services.

Watch our webinar on AI for local authorities and policymakers featuring  Guido Rink from the European Committee of the Regions and deputy mayor of Emmen

The first step

Investing in digital literacy and associated skills development is the first step to be ready for AI in the public sector. This includes raising awareness about the potential of AI solutions. Such measures create the conditions for the uptake of trustworthy solutions, leading to cultural and organisational change. Integrating training and capacity-building programmes for public servants is essential. 

Read our good practice AI Certificate for Public Servants from the Interreg Europe ENAIBLER project

In the uncertainties that AI technologies bring to the life of citizens and policy makers, the EU AI Act is a critical step forward, offering both immense potential for AI-driven innovation in local governments and highlighting the importance of vigilance against risks such as privacy breaches and biases. 

As the main actors on the ground, local authorities can rely on the new legislative framework to harness AI creating sustainable, inclusive, and resilient territories, ultimately improving the quality of life for citizens.

To go further

• Access our tailored expert services to create effective digital transformation policies embracing AI. Have a glance at our peer review on the Trustworthy use of AI in companies for the benefit of the Austria Wirtschaftsservice.
• Opinion paper on European Approach to Artificial Intelligence - Artificial Intelligence Act, Guido Rink
• Impact Assessment Mensenrechten en Algoritmes in Dutch, University of Utrecht
• Policy brief on Digital transformation of the public services