The European Parliament and the Council of the EU have given their approvals to the Gigabit Infrastructure Act (GIA) and on 8 May 2024 the GIA was published in the Official Journal of the European Union, with the Act entering in to force on 11 May 2024.

The GIA is aimed at reducing the costs and barriers to deploying next-generation gigabit broadband networks across member states, replacing its predecessor Broadband Cost Reduction Directive (BCRD) which was introduced in 2014. The BCRD was transposed into UK law as the Communications (Access to Infrastructure) Regulations 2016 (ATI Regulations). Suffice to say both incarnations didn’t quite work out as the policy makers planned yet their objectives remain as relevant as ever. The GIA has clearly been developed in an attempt to remedy the situation.

The stated objective of the BCRD, “to facilitate and incentivise the roll-out of high-speed electronic communications networks by lowering the costs of deployment with a set of harmonised measures”, was laudable, however the Directive was widely criticised and remains largely unutilised and ineffective due to its perceived shortcomings. The GIA, its much-anticipated replacement, while clearly aimed at addressing its predecessors’ issues, unfortunately retains much of the same language as the BCRD with a few notable tweaks here and there. For example, it retains an almost identically worded objective subject to swapping out the words “high-speed network” for “very high-capacity network” and the words “incentivise” for “stimulate” in relation to the roll out of high-speed networks. The exam question will therefore inevitably be: has the EU done enough to correct the flaws in the GIA’s predecessor?

A 2021 public consultation by the European Commission found stakeholders felt the BCRD had limited effectiveness in meeting its objective (of facilitating lower-cost, high-speed broadband deployment across the EU). Issues included persisting barriers to cost-effective rollouts, difficulties agreeing access terms and pricing, complex permitting procedures, and an absence of rules which would permit regulators to engage in robust enforcement. Similar reviews of the ATI Regulations in the UK echoed these concerns, unsurprisingly. With these foreseeable issues in the drafting, both the BCRD/ATI Regulations likely failed at the first hurdle of encouraging any rational access seeker/investor from developing (or trying to develop) a viable business case for network expansion based on access ‘rights’ under the BCRD.

The GIA aims to takes a more comprehensive approach[1] to reducing gigabit network deployment costs and addressing the BCRD’s shortcomings. We take a closer look into what’s changed and assess whether the new Act is likely to meet its objectives, this time, below.

Perhaps the most obvious and fundamental difference with the BCRD, is that, as an EU Regulation, the GIA will be directly effective across member states, unlike the BCRD which required national transposition as a Directive. This should lead to more consistent application and harmonisation across the EU at least and discourage forum shopping to some degree. This will be particularly important for aligning with the timing for when access seekers and providers rights and obligations come into force for strategic and investment planning. Ensuring greater pan-EU consistency is no doubt also expected to play a crucial role in pricing and dispute resolution and dovetails with the greater level of guidance given on that topic (more on that below). Of course, all of this aligns and is part of a piece within the EU’s latest big push on its single market aspirations.

A major flaw identified with the BCRD was the lack of clear pricing principles for granting network access, relying on a vague requirement that it must be given on “fair and reasonable” terms that allowed “a fair opportunity [for the access provider] to recover its costs, and to take into account the impact of access on the infrastructure operator’s business plan and investments.” This provided little legal certainty for access seekers, given the low incentives for network operators to provide access and/or offer reasonable rates, and so enabled significant leeway to refuse.

Unfortunately, the GIA still appears to contain a lot of the same flex for access providers to refuse access on the basis of subjective and potentially amorphous reasons such as undue disruption of “business plans” or “investments” and “the future need for space of the access provider”. Many of the bases that access requests may be refused have been largely copied across from the BCRD but for the addition of words such as that refusals must be “duly justified” – which one would think was implicit in the BCRD in any event.

The new Act has gone some way further, however, in providing clarifications on what is meant by “fair and reasonable” pricing for access requests. For instance, the GIA now requires that pricing must “reflect market conditions” and provides specific clarifications and criteria to be considered that require balancing a “fair return on investment” while “avoiding excessive prices”. The Recitals provide some useful steers e.g.,

“…certain existing contracts and commercial terms and conditions agreed between access seekers and access providers could be used by either access providers or dispute settlement bodies as a benchmarking factor to determine whether prices, and terms and conditions are fair and reasonable…”

The GIA also encourages Member States to provide guidance on the terms and conditions of access, including the price, in order to facilitate the conclusion of such agreements.

This contrasts with the BCRD’s failure to provide much at all by way of guidance or steer on what it meant by “fair and reasonable” in this context, but still set against a similar long list of rights of refusal, which arguably permitted wide scope for constructive denial of access and undermined legal certainty. In that light the GIA has made a few steps forward on pricing, but not giant leaps.

The BCRD focused mainly on access to “telecoms-like” utility network infrastructures e.g., pipes and cables. Given that price controlled passive access remedies, such as duct and pole access, were being introduced at greater rates across Europe by telecoms regulators soon after the BCRD was introduced this may have led to many access seekers asking themselves “Why bother with the hassle of trying to connect up to a gas pipe through the (flawed) BCRD”. Fast forward 10 years and needless to say a lot has changed, including advances in technology which makes new and innovative use of network topologies and ‘novel’ sites, which provides a greater role for a wider scope of infrastructure for access seekers to make use of.

The GIA appears aimed to recognise this by expanding the scope of the definition of “physical infrastructure” to cover accessing any physical infrastructure capable of hosting fibre, antennas, and network equipment. This aims to open up more up-to-date and viable routes for gigabit network deployment compared to the BCRD. For example, the definition of physical infrastructure has been extended to include street furniture, such as street lights, street signs, traffic lights, billboards, bus/tram stops, and metro stations[2].

The act also enables the possibility of operators accessing commercial buildings in rural regions where criteria are met, such as there being no suitable alternatives in order to deploy very high-capacity networks (VHCNs).

The approach under the BCRD to granting permits has been heavily criticised as enabling overly bureaucratic processes. While it did go as far as to require that all relevant information concerning the conditions and procedures applicable for granting permits should be available via a single information point (in an attempt to simplify permit processes involving multiple authorities) it did not provide details on the consequences of an authority failing to meet the deadline for permit approval or rejection. The BCRD only went as far as to merely suggest that Member States could consider measures such as tacit approval[3] or simplified procedures to ensure that decisions were made within the specified timeframe. No clear mechanism for operators to seek remedies in cases of permit-granting delays were specified as needing to be adopted.

To remedy this, the draft GIA originally included a proposal to implement a mandatory ‘tacit approval’ principle for permit applications for civil works. This meant that if a competent authority did not respond to a request for a grant of a permit within four months[4], it would be treated as if the authority had given tacit authorisation to proceed.

However, this proposal was changed in the final draft (some consider it ‘at the last minute’), with the EU opting instead for including the ability for Member States to decide to derogate from tacit approval in favour of a conciliation mechanism between public sector bodies and telecom operators[5]. This was apparently in response to objections from certain Member States, and was a means of diluting tacit approval, allowing EU countries to opt for alternative remedies, namely compensation mechanisms for delay[6] or the right to file complaints in court for permit delays. This conciliation mechanism nevertheless does serve as another attempt at an intermediary step aimed at expediting permit approvals.

Despite these changes, concerns have been raised by industry stakeholders regarding the reduction in measures intended to streamline timing and costs associated with network rollout. For example, the Association of European Telecommunications Network Operators emphasised the departure from the original ambition set by the European Commission, warning of potential fragmentation.

Like the BCRD, the GIA encourages shared infrastructure use, enables co-deployment during civil works, and includes measures aimed at simplifying rollout procedures across the EU. However, it also goes further by requiring that all new buildings or buildings subject to major works should be equipped with VHCN infrastructure[7] – a notable addition compared to the BCRD’s goal of in-building physical infrastructure needing to reach speeds of at least 30 Mbps. The GIA also places greater emphasis on reducing environmental impacts through efficient technologies (e.g., fibre and 5G), infrastructure re-use, and greater efforts in civil works coordination.

The real-world impacts of GIA will not be felt immediately upon its enactment. The regulation is set to come into force 18 months after its publication in the Official Journal, and there will be a staggered implementation of certain provisions, such as the digitalisation of single digital entry points, taking effect after 24 months. This delay allows Member States time to adapt to the new requirements.

Once the GIA is in force, Member States will need to amend or repeal any provisions that conflict with the regulation. However, as the GIA follows a ‘minimum harmonisation’ approach, Member States have the flexibility to introduce additional provisions that are stricter or more detailed than the minimum requirements set out in the regulation. This may lead to some inconsistencies in the application of the GIA across the EU, which may be further impacted where Member States derogate from tacit approval, creating a sliding scale of leniency.

Both the GIA and EECC both include measures aimed at facilitating the roll-out of very high-capacity networks, access to existing infrastructure, permit granting processes, and more. This overlapping scope could raise concerns about inconsistent approaches being taken and obligations being imposed on telecom operators and infrastructure owners creating legal uncertainty and disproportionate burdens.

Recital 21 clarifies the position however, that where access to certain infrastructure for a particular purpose is already available under the EECC, access via the GIA does not apply. However, where access is given by an NRA to certain infrastructure by reference to the EECC which includes a ‘usage restriction’ e.g., the EECC remedy is for wholesale dedicated capacity but not connecting mobile base stations, access via the GIA would be available. This is reflected in Article 3(5) of the GIA. As noted, the GIA also attempts to streamline permit granting procedures and limit fees, which complements Articles 43 and 16 of the EECC, respectively, as stated in Recitals 39 and 43 of the GIA.

Ultimately, the GIA acknowledges the primacy of the EECC in case of any conflicting provisions, stating in Article 1(2) that if any of its provisions conflict with the EECC, the relevant EECC provision shall prevail.

It appears that effort has been given to see that the GIA and EECC are complementary to create a harmonised framework which builds upon the existing provisions of the EECC. However, there still appears to be some potential pain points and opportunities for forum shopping where two access regimes are operating in parallel….with one requiring a finding of Significant Market Power before access obligations can be imposed, and the other does not. Reconciling issues and arguments that crop up will be key to maintaining efficient rollout, which has been deftly passed to the dispute resolution bodies of Member States to sort out.

On balance, the GIA heralds a significant improvement over the BCRD in harmonising the framework aimed at reducing deployment costs and laying a stronger foundation for more efficient gigabit ready network rollout. The two biggest improvements seem to be the implementation of the framework as a Regulation rather than a Directive, which should go a long way towards harmonisation across the EU and providing more detail and guidance to Member States around FRAND pricing principles.

However, some amendments on fundamental issues such as rights of refusal of access are arguably little more than tweaks around the edges and may not provide much more confidence or certainty for access seekers or investors alike. This could either discourage use of the access framework yet again or result in years of litigation to bottom it out.

In the UK, government has sporadically made a show of revisiting the equally failed UK local version of the ATI Regulations.[8] However, with the success of Ofcom’s communications sector specific, regulated cost based price controlled Physical Infrastructure Access remedy imposed on Openreach (and its ongoing £5 billion Project Gigabit subsidy scheme acting in parallel) there may be little appetite for the UK government to be quick to follow the EU and update the ATI Regulations. Especially with an election on the horizon…but watch this space.

Nonetheless, the GIA marks an important step towards the EU’s connectivity and harmonisation goals, aiming to address the key criticisms of the BCRD. With the GIA now adopted by EU policymakers, all eyes turn to how (and when) the mechanisms will work in practice. In this fast-changing environment, there is also the small matter of how much demand there will actually be from industry to use it.

References

[1] Gigabit Infrastructure Act – https://digital-strategy.ec.europa.eu/en/policies/gigabit-infrastructure-act#:~:text=A%20political%20agreement%20was%20reached,complex%20procedures%20for%20network%20deployment

[2] The purpose of which is to support small cells deployment for 5G.

[3] Meaning that no response from a regulator was to be interpreted as acceptance.

[4] While both pieces GIA and the BCRD allow for extensions to this procedure in exceptional circumstances, the GIA seeks to cap the extension to a further four months. The BCRD on the other hand didn’t provide such a cap, and enabled discrepancies as to the term permitted, stating any extension should “be the shortest possible in order to grant or refuse the permit”.

[5] Note the tacit approval method remains the default under the directive.

[6] The maximum duration for granting permits is four months under the GIA.

[7] The current European gigabit connectivity goal is 1 Gbps.

[8] https://www.gov.uk/government/publications/review-of-the-access-to-infrastructure-regulations-call-for-evidence/review-of-the-access-to-infrastructure-regulations-call-for-evidence#ATI-4-6