BEREC high-level position on the interplay between telecoms & connectivity and Artificial Intelligence and Virtual Worlds
Written By
The Body of European Regulators for Electronic Communications (BEREC) has released a comprehensive report examining the implications of Artificial Intelligence (AI) in the telecommunications sector, recognising that connectivity is crucial to support AI and Virtual Worlds (VWs). These technologies hinge on the successful deployment of high-quality Electronic Communications Networks and Services (ECN/ECS), with technologies such as cloud and edge computing playing a pivotal role.
In the context of competition dynamics, BEREC identifies four essential inputs for AI: Computing Power, Data, Financial Resources, and Technical Expertise. Exclusive access to these inputs by certain market participants can create significant competitive advantages and establish structural barriers for newcomers. In this context BEREC notes that regulatory tools, such as the Data Act, the Digital Markets Act (DMA), and the AI Act, could influence the developing landscape for AI technologies. These laws could help promote the development of a dynamic ecosystem around the four essential AI inputs.
Regarding internet openness (the way content is accessed and distributed online) and end-user empowerment, BEREC emphasises that end-users are not only beneficiaries of AI services but also contribute to their development by providing data and feedback. AI's potential to enhance user experience and alter traditional methods of accessing online content and services is significant and could impact access to content (notwithstanding the underlying net neutrality framework which targets internet access providers). However, as a new intermediary, AI can influence user access and potentially impact user choice. It is therefore critical to assess AI's potential impact on internet openness. To mitigate such risks, BEREC emphasises the importance of transparency and boosting end-users' control, understanding, and rights. This includes enhancing transparency and comprehensibility, strengthening consent and control frameworks, promoting data portability and access, implementing privacy enhancements, and combating fraud, scams, and illegal behaviours.
BEREC also considers the cyber risks posed by AI on ECN/ECS, as well as the solutions which AI can provide to address such risks. The risks identified include the use of large language models (LLMs) to automate malicious software code or produce fraudulent text, speech and videos with a level of credibility which makes detection difficult. This also extends to VWs where the cyber risks may arise in the form of virtual data and property theft, or the hacking of the devices providing access to VWs themselves. However, BEREC also identifies that AI can be utilised to address such risks by improving real-time detection of fraudulent behaviour and the elimination of threats prior to occurrence.
In terms of environmental footprint and sustainability, BEREC highlights the importance of minimising the environmental impact of digital technologies, including AI systems and VWs, in order to achieve the EU's targets for reducing greenhouse gas emissions. These systems require significant resources, including water, energy, and materials for servers and equipment housed in data centres, leading to various environmental impacts and sustainability challenges. Accurate assessment of these impacts is necessary, necessitating robust data collection and sustainability indicators.
BEREC also recognises that AI-based services could potentially enable more sustainable solutions across various markets. For example, AI can optimise individual and collective electricity consumption using data from smart monitoring devices. However, the sustainability of the AI systems themselves is crucial for this potential contribution to green transition to be effective.
Environmental studies are required to identify strategies to reduce AI's environmental footprint towards "frugal AI" and sustainable VWs. Environmental considerations should be incorporated in EU policies that support innovation, and efforts should be made to minimise the impact of digital services on the footprint of devices and infrastructure by promoting proper eco-design criteria and environmental transparency.
In summary, the BEREC report provides a detailed examination of the role of AI in the telecommunications sector, highlighting the importance of addressing competition dynamics, internet openness, cybersecurity risks and environmental footprint. It underscores the importance of robust legislation to shape the regulatory environment and promote competition, the critical role of end-users and the need for greater empowerment and transparency, and the need to consider cyber risks and environmental sustainability in the development and deployment of AI systems.
For more information, please contact Max Johnson and Anthony Rosen.
Regarding internet openness (the way content is accessed and distributed online) and end-user empowerment, BEREC emphasises that end-users are not only beneficiaries of AI services but also contribute to their development by providing data and feedback. AI's potential to enhance user experience and alter traditional methods of accessing online content and services is significant and could impact access to content (notwithstanding the underlying net neutrality framework which targets internet access providers). However, as a new intermediary, AI can influence user access and potentially impact user choice. It is therefore critical to assess AI's potential impact on internet openness. To mitigate such risks, BEREC emphasises the importance of transparency and boosting end-users' control, understanding, and rights. This includes enhancing transparency and comprehensibility, strengthening consent and control frameworks, promoting data portability and access, implementing privacy enhancements, and combating fraud, scams, and illegal behaviours.
BEREC also considers the cyber risks posed by AI on ECN/ECS, as well as the solutions which AI can provide to address such risks. The risks identified include the use of large language models (LLMs) to automate malicious software code or produce fraudulent text, speech and videos with a level of credibility which makes detection difficult. This also extends to VWs where the cyber risks may arise in the form of virtual data and property theft, or the hacking of the devices providing access to VWs themselves. However, BEREC also identifies that AI can be utilised to address such risks by improving real-time detection of fraudulent behaviour and the elimination of threats prior to occurrence.
In terms of environmental footprint and sustainability, BEREC highlights the importance of minimising the environmental impact of digital technologies, including AI systems and VWs, in order to achieve the EU's targets for reducing greenhouse gas emissions. These systems require significant resources, including water, energy, and materials for servers and equipment housed in data centres, leading to various environmental impacts and sustainability challenges. Accurate assessment of these impacts is necessary, necessitating robust data collection and sustainability indicators.
BEREC also recognises that AI-based services could potentially enable more sustainable solutions across various markets. For example, AI can optimise individual and collective electricity consumption using data from smart monitoring devices. However, the sustainability of the AI systems themselves is crucial for this potential contribution to green transition to be effective.
Environmental studies are required to identify strategies to reduce AI's environmental footprint towards "frugal AI" and sustainable VWs. Environmental considerations should be incorporated in EU policies that support innovation, and efforts should be made to minimise the impact of digital services on the footprint of devices and infrastructure by promoting proper eco-design criteria and environmental transparency.
In summary, the BEREC report provides a detailed examination of the role of AI in the telecommunications sector, highlighting the importance of addressing competition dynamics, internet openness, cybersecurity risks and environmental footprint. It underscores the importance of robust legislation to shape the regulatory environment and promote competition, the critical role of end-users and the need for greater empowerment and transparency, and the need to consider cyber risks and environmental sustainability in the development and deployment of AI systems.
For more information, please contact Max Johnson and Anthony Rosen.
