"Only four of the world's top 50 tech companies are European," Mario Draghi told the European Parliament in Strasbourg on 17 September 2024.
The telecoms chapter of his report was one attempt to answer that sentence. Europe's networks are built and regulated 27 times over. Spectrum is licensed nationally on different terms and different clocks. Operators are numerous and sub-scale. The report treated all of this as a single problem with a single consequence: not enough capital reaches the infrastructure that everything else runs on.
The Commission's answer is the Digital Networks Act, tabled on 21 January 2026. EPIC, the Brussels-based European Policy Innovation Council, has been grading it — and the grades are worth reading before the trilogue starts moving numbers around.
EPIC's Draghi Observatory, published at draghiwatch.eu, tracks how many of the report's 383 recommendations have become binding EU law. Alongside the main index it runs a series of spin-off briefs under the title "From Draghi to DNA", looking specifically at the telecoms file.
The first of those, in November 2025, was a feasibility assessment written before the Commission's text existed. It took the 20 telecoms recommendations in the Draghi Report and asked a narrow question: how many of them could a Digital Networks Act actually deliver? The answer was about 35%.
That is not a criticism of the Commission's drafting. It is a statement about instrument choice. Several of Draghi's telecoms recommendations are not really telecoms regulation. Cross-border consolidation runs through merger control and competition policy. Spectrum assignment sits with member states, which treat auction revenue as national money. A regulation on electronic communications can adjust the conditions around those questions; it cannot decide them. If two-thirds of a recommendation set lies outside the instrument chosen to implement it, the delivery ceiling is fixed on the day the instrument is chosen.
The second brief, an alignment scorecard published in February 2026, did the harder job: it read the Commission's actual legal text against all 20 recommendations. Nine were judged aligned, seven partially aligned, four not aligned.
Read the two briefs together and a sequence appears. EPIC's feasibility estimate said roughly seven of the 20 were deliverable in this instrument. The legal text, once written, aligned fully with nine. On EPIC's own numbers the Commission drafted slightly above the ceiling its analysts had set — which is a better result than the headline "fewer than half" suggests, and worth saying plainly.
The scorecard also captures the text at its most ambitious moment. A Commission proposal is the high-water mark of a file. What follows is Parliament and Council, and both are already reshaping it.
The Act would replace much of the existing telecoms rulebook. On spectrum, licences of indefinite duration become the principle, with a minimum term of 40 years where a limited licence can be justified, plus spectrum sharing and "use it or share it" obligations — a direct attempt at the investment-predictability problem Draghi identified. On networks, it mandates the copper switch-off: by 31 December 2035 everywhere, and earlier in areas that pass two cumulative tests, 95% fibre coverage measured in homes passed and the availability of affordable retail connectivity. It also covers satellite authorisation, network resilience and cybersecurity, and cross-border market access.
The Council is pushing for EU-wide benchmarks defining "affordable", to guarantee a floor of protection for vulnerable consumers that member states may improve on. Spectrum and satellite authorisation are widely expected to be the most contested items in the trilogue, which is to say the provisions closest to Draghi's diagnosis are the ones most likely to be diluted.
Adoption is expected in late 2026 or 2027. Application is not expected before the end of 2027 — a little over three years after the speech in Strasbourg.
The telecoms file is a specific case of a general result. EPIC's July 2026 preliminary update to the Draghi Implementation Index put strict implementation at 15.7% of the 383 recommendations, or 60 of them, and strict-plus-partial at 41.3%, or 158. The February-to-June half-year added 2.4 percentage points against 7.5 in the preceding one: delivery is decelerating.
The same update produced EPIC's first ranking by directorate-general. DG TRADE led on 41.7% strict implementation. DG ENER managed 2.7%. DG EMPL recorded zero. The Observatory's reading is consistent across files: the EU moves fastest where competitiveness fuses with security, and slowest where delivery means forcing a market outcome that some member state currently owns.
Telecoms sits squarely in the second category. National spectrum, national incumbents, national revenue.
An index built on one report's recommendations measures fidelity to that report, not the quality of the law. The Commission never promised that the Digital Networks Act would implement all 20 of Draghi's telecoms recommendations, and a scorecard that grades it as though it had is measuring against a standard the drafters did not accept. It is also possible to think some of the four unaligned recommendations were the right ones to leave out.
What the scorecard does establish is narrower and still useful: on the one file where the Commission chose to legislate directly against a Draghi chapter, the proposal covers under half the recommendations in full, and the parts most likely to be traded away in negotiation are the structural ones.
Europe's telecoms problem was never that nobody had written it down. It was written down in September 2024, costed, and handed to the institutions with a list attached. Two years on, the instrument chosen to answer the list can only reach about a third of it, the text answers rather more than that, and the process that follows can only subtract.
EPIC's next full review of the Draghi Index lands in September 2026, the report's one-year-of-tracking mark. The Digital Networks Act will not be law by then. When it is, the useful question will not be whether it implemented Draghi. It will be whether the parts that survived the trilogue were the parts that moved capital — or the parts that nobody in a national capital had a reason to defend.
