Artificial Intelligence and virtual worlds: BEREC adopts a high-level position

BEREC submitted its high-level position on Artificial Intelligence (AI) and virtual worlds (VW) to the European Commission. The high-level position provides a holistic perspective and focuses  on a variety of aspects building on BEREC’s broader expertise. 

BEREC’s position builds on previous work, such as BEREC reports on the application of AI solutions for the provision of ECN/ECS, on the internet ecosystem and the ex-ante regulation of digital gatekeepers.  

In order to unleash the potential of AI and VW, and to make sure that these technologies develop for the benefit of European citizens, BEREC believes that some key issues concerning their competition dynamics, internet openness, environmental footprint and sustainability, and cybersecurity need to be assessed and tackled.

Good ECN/ECS are needed for AI and VW to thrive
In the adopted document, BEREC indicates that, for AI and virtual worlds to thrive, there is a need for good ECN/ECS, such as very high-capacity networks with low latency, as well as for high availability of cloud and edge computing.  

Healthy competition dynamics
BEREC emphasises that integrating AI technologies can lead to transformative benefits in innovation, efficiency and service quality. However, without careful supervision, this could also exacerbate competition issues, potentially harming users’ choice and innovation. Indeed, some major actors have privileged or exclusive access to the few key inputs on which AI relies and may have the incentive to impose unfair terms and conditions on other firms or users.

Human-centric approach 
The high-level position also states that the end-users play a crucial role in the development of AI as they provide data and feedback, which helps to improve and further develop AI services. However, by shaping the way users access content online, services embedding generative AI can potentially affect users’ freedom of choice. Moreover, the increasing pervasiveness of AI raises important questions about internet openness, privacy, cybersecurity, data integrity, reliability and societal biases perpetuated by algorithmic bias and complexity. BEREC suggests adopting a collection of possible solutions – a regulatory toolbox that should include specific tools to enhance the principle of a human centric approach to generative AI and VW. 

Environmental sustainability and security perspective
BEREC also touches upon the impact of AI and VW on sustainability and emphasises that, like for other digital technologies, AI’s environmental footprint should be properly assessed to minimise adverse effects and promote the potential use cases that can benefit the twin green and digital transitions. AI-based services’ contribution to the green transition can only be effective if AI systems are themselves sustainable.

With regard to cybersecurity, BEREC emphasises the need to focus on building a strong and secure foundation that can effectively manage risks and take network security to the next level. 

BEREC contribution to the regulation of digital markets
BEREC will keep analysing the evolution of digital markets and will continue to collaborate with the EU institutions and other European bodies and networks to identify potential challenges and contribute to the definition of possible solutions, and to ensure that new digital technologies, services and products are provided for the benefit of European citizens. 

BEREC already contributes to the Digital Markets Act (DMA) High-Level Group, which provides the European Commission with advice, recommendations and expertise for any general implementation of the DMA to ensure that EU digital markets are contestable and fair. Additionally, in the context of the Data Act, several BEREC members may be designated as the competent authority implementing some chapters relevant to AI, such as switching obligations between data processing services and interoperability, among others. 
In the future, BEREC members could also play a role in implementing the AI Act at national level, as the national regulatory authorities should be equipped to address sectoral competition concerns regarding AI applications.