BEREC believes that existing regulation together could mitigate many of the risks AI poses to online users, but is mindful that gaps in the EU framework may remain
Generative AI is creating a new layer of intermediation within the internet ecosystem
On 1 October 2026, the Body of European Regulators for Electronic Communications (BEREC) published its draft report on the impact of generative AI on internet openness. BEREC highlighted that generative AI is creating a new layer of intermediation within the internet ecosystem through its integration into online services, including search engines, messaging platforms and operating systems. While the body acknowledged that the adoption of AI could benefit consumers and businesses by providing a more personalised experience, it also noted that the technology could limit their freedom of choice and reduce transparency. In particular, when using generative AI services, users delegate their choices to AI models, allowing them to select sources and summarise them on their behalf, which BEREC noted can be especially problematic when the model is inaccurate.
A shift from search engines to response engines may make end users less likely to access original sources themselves
BEREC outlined that AI has changed the way users access online content, with searches shifting from traditional search engines to response engines, which generate a synthesised answer and provide a limited number of sources. The body cited research suggesting that this comes with a trade-off, where users accept lower transparency in exchange for access to personalised content, and that the default integration of generative AI features into digital services can limit users’ ability to control their use. BEREC highlighted that users of Google are less likely to click on the link to a source when the firm’s ‘AI Overviews’ provides the answer to their query, resulting in a ‘zero-click’ interaction. This reduces traffic to third-party sites, lowering the revenue these sites can earn from advertising or subscriptions. In turn, such interactions undermine the economic incentives to produce human-generated content, despite this content remaining essential for the quality of generative AI systems themselves.
Generative AI services have also led to a proliferation of synthetic content on the internet, including news and audiovisual platforms, which BEREC notes can make it more difficult to verify sources and assess the quality of online information. The body suggests that this could create a “structural paradox”, whereby AI systems depend on continuous access to high-quality online content while simultaneously weakening the economic incentives to produce such content.
No single piece of regulation currently addresses these challenges, but effective coordination of several frameworks could mitigate risks
BEREC considered that, while no single European regulatory instrument can fully address the broader impacts of generative AI on an open internet, several frameworks can tackle some of the challenges posed by the technology. For example, the Digital Markets Act (DMA) contains rules to prohibit some problematic gatekeeper behaviours such as self-preferencing, while the Digital Services Act (DSA) provides a framework to identify and mitigate systemic risks affecting media pluralism and access to information. The AI Act also imposes obligations relating to transparency, which could help users to understand when they are relying on AI-generated information or decisions. BEREC suggested that effective coordination of these rules will be key for preserving internet openness, and that priority should be given to the implementation of the existing framework, particularly as certain pieces of regulation have only recently become applicable. However, the body noted that further monitoring will be needed to determine whether any regulatory gaps remain.
