The European telecoms regulatory framework: not a good fit for the public cloud

This paper contributes to the question, raised by the European Commission (EC) in its recent white paper on the future of digital infrastructure in Europe, of whether the cloud and telecoms sectors may be converging, to the extent that common regulation would be justified. Specifically, the EC outlines the option to expand the European Union (EU)’s telecoms regulatory framework to include cloud services. In this paper, we examine this question from a technical, legal and economic perspective, considering the history of the telecoms sector and the purpose for which the telecoms regulatory framework was constructed and implemented.

Cloud services allow European businesses to access IT building blocks running over distributed infrastructure. Public-cloud services are designed to be useable across industries, through common application programming interfaces (APIs). These services are underpinned by infrastructure that is distributed globally and connected via extensive private network links. European businesses benefit from cloud services financially, because they can access extensive IT resources with limited up-front investment and risk. They benefit operationally because they can access state-of-the-art IT building blocks, which very few businesses may have been able to source and access in a dedicated manner. 

Businesses use cloud services through many independent software vendors (ISVs) which offer software on cloud platforms. This includes telecoms operators, which use cloud-based services offered by a range of vendors, most of which had been offering on-premises software for decades. Telecoms operators have begun migrating some of their non-network IT to public-cloud platforms, but migration of network IT remains very limited (less than 1% of workloads by some estimates), with no clear momentum towards greater use of the public cloud for network functions. The claims of ‘convergence’ are therefore at best premature, and at present largely inaccurate. Cloud providers and customers are indeed dependent on connectivity to be able to work together, but telecoms operators are likely to remain largely independent from cloud providers in the context of running their network. As they migrate network functions to the public cloud, they will do so using software-defined networking solutions provided by vendors such as Nokia and Ericsson, building on the same cloud services as are available to all other businesses.

Telecoms regulation (now under the European Electronic Communications Code, EECC) reflects a history of state-controlled monopolies, and the policy decision that regulation should support market liberalisation and competition. This translated into a strongly pro-competition ex-ante regulatory regime that required national regulatory authorities (NRAs) to review specific relevant markets and impose remedies on operators with significant market power, in addition to general conditions of authorisation. Interconnection between telecoms operators was and remains subject to regulation, reflecting the importance of direct network effects in traditional telecoms markets, in particular telephony.

By contrast, the cloud sector is relatively new, highly innovative and dynamic, with many providers competing for customers in different ways. Direct network effects are largely absent, but economies of scale are strong and not bound by national borders. The sector is already overseen through European competition law, and has recently been brought under the scope of new regulations including the Data Act, the Digital Market Act (for the largest providers), the Digital Services Act, and the revised Network and Information Security Directive (NIS2). Indeed, competition authorities have taken an interest in the competitive dynamics related to cloud services, and highlighted some concerns related to egress fees, barriers to switching and software licensing practices. If any regulatory concern is identified after testing these new instruments, regulators should seek to remedy them through proportionate and justified remedies, subject to a detailed impact assessment: the EECC was not constructed for this purpose and appears highly unlikely to be effective, justified and proportionate in addressing these potential remaining issues.

If the EC chooses to expand regulation to cloud services, it should conduct a detailed impact assessment. In the last section of the paper, we outline potential impacts for European cloud and telecoms providers, and end users in both sectors. We find that European cloud providers may face higher costs and reduced incentives for investments in Europe. Competition in the telecoms sector may be distorted in favour of larger operators, which have championed the regulation of IP interconnection as a way to extract payments to terminate internet traffic to their subscribers. Eventually, these effects would harm European businesses, affecting their ability to adopt, and benefit from, cloud and artificial intelligence (AI) services, which would be counterproductive to Europe’s digital agenda and its ability to innovate through technology.

In conclusion, we reiterate the importance of well-functioning cloud and telecoms sectors to the digital agenda for Europe, and to the European businesses and public-sector organisations that use cloud services and stand to benefit from them, including in the context of AI and other highly innovative aspects of IT and digital technology. This is essential to Europe’s competitiveness. Regulators should acknowledge the potential adverse impacts of extending the telecoms regulatory framework to encompass cloud services, without clear justification or assessment of its impacts. A nuanced approach, recognising the unique characteristics and dynamics of both sectors, is essential to avoid these risks and support continued growth and innovation for European businesses.