Summary

  • In BEREC’s 2025–26 survey of 28 national regulators, 25 sought provider information, 24 used market surveys, 24 analysed complaints and user reports, and nine used technical network monitoring to inspect internet-access services’ commercial and technical conditions.
  • The report asks separate questions about traffic management, contract transparency and high-speed access. Their different counts describe different oversight jobs. It also records France’s 136 relevant platform alerts, Portugal’s roughly 720 direct internet-access complaints, and Poland’s 57 completed certified speed measurements.

The desk and the socket

Imagine a regulator trying to find out whether an internet provider treats traffic as the rules require. The provider can send a reply. A customer can send a complaint. An official can read the offer on the provider’s website. A test can inspect what the connection actually does. These are four doors into the same building; they do not open onto the same room.

BEREC’s tenth implementation report makes the choice unusually visible. It asked 28 national regulatory authorities, through 27 questions, about work from 1 May 2025 to 30 April 2026. On the question of monitoring the commercial and technical conditions of internet-access service, 25 used information requests to providers. Twenty-four checked the market without asking providers, such as by reading offers. Twenty-four examined complaints or end-user reports. Nine reported technical network monitoring. One cannot tell the state of Europe’s packets from the thickness of a regulator’s correspondence file.

The figures describe methods used, not cases won or operators found at fault. Their value is architectural. A contract audit can catch a promise that misstates speed. A complaint can identify a fault that hurts a person. A network test can catch behaviour that does not arrive pre-labelled as a complaint, provided the test is designed for it. Each method has a blind side. A serious watchdog needs to know where its own eyes are.

What customers bring to the counter

The complaint channel is real, and the report puts numbers on its shape. Twenty-three of the 28 authorities collect or monitor complaints concerning the end-user rights and transparency provisions addressed in its questionnaire. The issues they receive are often the ones a customer can feel: slow service, disconnection, a bill or a contract. That is useful information. It is also an agenda set partly by what people can recognize and describe.

France reported 136 alerts concerning these open-internet obligations on its J’alerte l’Arcep platform during the period, out of more than 66,500 alerts on the platform overall. Those 136 were reports from users, not formal complaints. Portugal reported roughly 720 direct complaints about internet access, a 99% increase on the previous reporting period. In the descriptions of the Portuguese complaints, 74% mentioned fixed-service faults, 14% mobile faults, 8% fixed speeds below advertised or contracted levels, and 2% traffic shaping. Those are the subjects people reported; a regulator still has to establish what occurred.

The contrast is practical. A broken fixed connection is hard to ignore. Preferential treatment of one class of traffic may be less obvious to a person watching a spinning wheel. If the public channel principally supplies the former, the regulator needs another route to inspect the latter. Otherwise the most legible failure becomes the most governable failure, and the quiet one gets to wear an invisibility cloak.

The test bench exists, but it answers several questions

BEREC did not ask a single yes-or-no question labelled “does this regulator measure the internet?” Its different chapters count different work. For oversight of traffic-management practices, 21 authorities used user complaints or reports and 21 requested provider information; 15 used market surveys. For transparency duties under Article 4, 21 requested information and 19 each used market surveys and end-user material. For availability of high-speed internet access under Article 5, 22 authorities monitored it; 17 used user reports, 13 provider requests and 12 technical network monitoring. The 12 and the nine come from different questions.

Swapping them would make a tidy chart and a bad article.

Other measurement routes appear as well. Twenty of the 28 authorities offered a quality-monitoring mechanism to consumers. Poland’s regulator reports that 64 certified measurement kits were loaned between 25 July 2025 and 30 April 2026, with 57 measurements completed. Seven download results and 12 upload results fell below 80% of the typically available speed declared by the applicants. The counts are for directions of measurement, not unique guilty connections. They show a path from a person’s speed concern to a reproducible instrument reading.

Measurement is therefore present. The governance question is how its results join the rest of the case. Who chose the sites and times? Which applications and traffic paths were tested? Was the result repeatable? Did a provider’s explanation survive a test it did not design? BEREC’s report maps national activity; it does not publish a common case-by-case trail connecting each method to a finding. That missing join is where an impressive monitoring inventory can still leave a reader guessing.

A rule needs a way to know

The open-internet rule has spent ten years in force. BEREC’s report says authorities have developed consistent and harmonised implementation over that span. It also records that 27 of the 28 respondents reported no change in their approach to monitoring commercial and technical access conditions since the previous period. A stable method can be a sign of an established practice. It can also preserve a stable blind spot. The report’s method counts are a reason to ask for the results of each route, not to award a medal for how many forms an office can send.

An independent regulator has the authority to ask, inspect, decide and enforce. The provider controls much of the information requested. The user controls the complaint they know how to make. The authority controls the investigation that converts either signal into a finding. That division of labour makes the measuring instrument politically important without making it glamorous. A quiet probe in the right place can be more useful than a loud press release about vigilance.

Sources