The European Commission’s 44th annual trade defence report puts 232 measures in force at the end of 2025, covering roughly 637,000 manufacturing jobs, and reveals a surveillance apparatus built to catch trade diverted by other countries’ tariff walls.
BRUSSELS, September 20, 2026
The European Union ended 2025 with 232 trade defence measures in force, opened 32 fresh investigations during the year and imposed 26 definitive duties, according to the European Commission’s annual accounting of its trade defence activity, published in Brussels on Friday, September 18.
The 44th Annual Trade Defence Report, adopted by the Commission on September 17 and released publicly the following day alongside an accompanying staff working document, describes a year in which demand for protection from European producers stayed close to record levels and in which the Commission built new machinery to detect goods being redirected into Europe by tariffs imposed elsewhere.
The Commission estimates that the measures in force at the end of 2025 protect around 637,000 manufacturing jobs across the bloc. It said the 32 initiations were just slightly below the record high set in 2024, and that demand reflects an ongoing increase in unfair trade practices harming European Union industry, remaining particularly pronounced in the chemical industry and in the iron and steel sector.
For exporters shipping into Europe, and for importers who buy from them, the report is the closest thing available to a map of where the European Union’s defensive perimeter currently sits and where it is likely to expand next.
The headline numbers
The report’s arithmetic repays close reading.
Of the 232 measures in force at the close of 2025, 220 were definitive, with the remainder provisional or otherwise interim in character. The Commission imposed provisional anti-dumping duties in 27 cases during the year. It imposed no provisional anti-subsidy measures at all in 2025, a figure that sits oddly against the political prominence of subsidy questions in European debate over Chinese industrial policy, and one that reflects the greater evidentiary burden and longer timelines that anti-subsidy work carries.
The Commission also applied measures retroactively in three cases, backdating duties to a period before provisional measures were introduced. Retroactivity is a rarely used power. It requires a finding that importers were aware, or should have been aware, of dumping and of the likelihood of injury, and that a surge of imports occurred before measures were imposed in a way that would undermine their remedial effect. The Commission presented the three cases as evidence of its willingness to call upon the trade defence instruments at its disposal, which is a diplomatic way of saying that stockpiling ahead of an expected duty is no longer a reliably safe strategy.
For comparison, the figure of 232 measures in force represents a defensive estate that has grown steadily for a decade. It spans anti-dumping duties, countervailing duties and safeguards, and it is heavily concentrated in intermediate industrial goods: steel and steel products, chemicals, fertilisers, ceramics, fasteners, glass fibre and aluminium. These are inputs rather than consumer goods, which means their cost effects propagate through European manufacturing rather than appearing directly on retail shelves.
The surveillance turn
The most consequential passage in the report is not a number but a description of process.
In June 2025, the Commission announced the launch of an import surveillance task force to monitor harmful trade diversion, meaning goods that might be redirected to the European Union because of high tariffs imposed elsewhere. The task force tracks imports and identifies potentially damaging flows. In conjunction with industry, the Commission uses that data to identify risks and, if necessary, to take appropriate action.
This is a structural change in how European trade defence begins. The classical model is reactive. A European producer or an industry association files a complaint supported by evidence of dumping, injury and causation, and the Commission decides within 45 days whether to initiate. The surveillance model inverts the sequence. Import data is monitored continuously, anomalies are flagged, and the Commission and industry discuss whether the numbers support a case before a complaint is drafted.
The practical effect is speed. A complaint built on data the Commission has already been watching is a complaint that can be assessed quickly and initiated without a long evidentiary courtship. That compression is visible in this year’s caseload. The safeguard investigation into grain-oriented electrical steel was initiated on March 27, 2026 and produced provisional measures on September 18, 2026, a turnaround of under six months for an instrument that is entitled to take nine.
The report also records a second piece of new machinery. During 2025 the Commission developed an exporter reporting and consignment tracking tool, which became effective at the start of 2026 to monitor a price undertaking concerning imports of battery electric vehicles. Price undertakings, in which exporters commit to sell at or above an agreed minimum price in lieu of paying duties, have historically been difficult to police because compliance depends on transaction-level data held by the exporter. A dedicated tracking tool suggests the Commission intends to make undertakings a more usable instrument, which matters well beyond the automotive file.
Enforcement, circumvention and the steel file
The report devotes considerable attention to enforcement of measures already in place, framing it as a response to increasingly complex attempts at circumvention.
Two steel actions from 2025 anchor that discussion. On March 25, 2025, the Commission adjusted the functioning of the European Union’s steel safeguard to preserve the measure’s effectiveness, tightening quota management in response to import patterns that were eroding its practical bite. On November 18, 2025, it imposed definitive safeguard measures on certain ferro-alloys to protect European industry from a harmful increase in imports.
Both decisions were staging posts on the road to the far larger intervention that followed. Regulation (EU) 2026/1384, published in the Official Journal on June 24, 2026, entered into force the next day and applied from July 1, replacing the expiring steel safeguard with a permanent regime. It reduced the total volume of duty-free import quotas by approximately 47 percent against the 2024 reference system and raised the out-of-quota duty to 50 percent. On August 31, 2026, the Commission adopted an implementing act specifying the evidence importers must provide to prove the country of melt and pour for products covered by the regulation, a rule aimed directly at origin transformation through third countries.
Read together, the annual report and the 2026 steel regime describe a single policy direction. Measures are being made harder to circumvent, quotas are being cut, out-of-quota duties are being raised, and origin is being verified at the point of melt rather than the point of last substantial transformation.
Defending European exporters abroad
The report is not solely an account of measures imposed. It also records work done on behalf of European exporters facing trade defence investigations in third countries.
The Commission said its services continued to assist European Union exporters subject to investigations abroad, citing the anti-dumping and anti-subsidy investigations opened by China into imports of brandy, pork and dairy products. Those three cases have been widely understood in Brussels and Beijing alike as responses to the European Union’s own countervailing duties on battery electric vehicles from China, and they target producer regions with acute political sensitivity in France, Spain, the Netherlands and Ireland.
This reciprocal dimension is the part of the trade defence story that is easiest to lose in the numbers. Every measure the European Union imposes creates an incentive for a trading partner to open a case of its own, and the sectors chosen for retaliation are rarely the sectors that caused the original dispute. Cognac producers did not build electric vehicles. That asymmetry is the point.
What it means for exporters shipping into Europe
For exporters, the report functions as a risk register.
The sectoral concentration is the first signal. Chemicals and iron and steel accounted for a disproportionate share of 2025 initiations, and both sectors continue to generate cases in 2026. A producer of an intermediate chemical or a steel product with rising European market share and falling European prices should assume it is visible to the surveillance task force, whether or not any complaint has been filed.
The retroactivity finding is the second. Three retroactive applications in a single year is enough to change the calculus on pre-duty shipments. The traditional response to an initiation, which is to accelerate shipments while the investigation runs, now carries a tail risk of duties being applied to goods that have already cleared customs and been sold. Importers who advance-buy into an open investigation should be pricing that contingency, and contracts should allocate it explicitly.
The undertaking tracking tool is the third. Exporters who negotiate minimum price undertakings should expect transaction-level reporting obligations and should build the data systems to meet them before agreeing to terms. An undertaking that is breached, even through administrative failure, converts into duties, and the conversion is not gentle.
What it means for European buyers
The mirror image of 637,000 protected jobs is a cost borne somewhere in the European economy.
Trade defence measures on intermediate goods raise input costs for the European manufacturers who buy them. A duty on a chemical intermediate is a cost to the formulator downstream. A safeguard on electrical steel is a cost to the transformer builder. The European Union’s own steel value chain has been unusually vocal about this in 2026, with downstream associations warning Brussels in early September about a widening competitiveness gap between protected upstream producers and exposed downstream fabricators who compete against imported finished goods that carry no equivalent protection.
That tension is the defining unresolved question in European trade policy. Protection applied at the raw material stage without corresponding protection at the finished goods stage transfers competitive advantage to foreign fabricators, who buy the same raw material at world prices and sell the finished product into Europe. The Commission’s decision this month to extend the electrical steel safeguard through laminations, cores and cores inside assembled transformers is the clearest indication yet that it has heard the argument and is willing to act on it, at the cost of considerable administrative complexity.
Buyers should also note the duration effect. Anti-dumping and anti-subsidy measures run for five years and are renewed following expiry reviews that, in practice, more often than not result in extension. The European Commission initiated an expiry review on September 16, 2026 into anti-dumping duties on cold-rolled stainless steel sheets and coils from India and Indonesia, following a request from the European steel association EUROFER on June 29, 2026. Duties in that case currently run at 10.0 and 35.3 percent for India and at 9.3, 19.3 and 20.2 percent for Indonesia. Until the review concludes, which the Commission says will be within 12 months, the duties remain in force. A measure notionally due to expire therefore does not expire on schedule, and sourcing plans built around a lapse date are built on sand.
The 2026 caseload so far
The report covers 2025, but the year since tells the reader where the trend is running.
The Commission imposed definitive anti-dumping duties in August 2026 on cold-rolled flat steel from India, Japan, Taiwan, Turkey and Vietnam. It has separately imposed definitive duties on hot-rolled flat products from Egypt, Japan and Vietnam at rates reported at 11.7 percent for Egypt, a range from 6.9 to 30 percent for Japan and 12.1 percent for Vietnam, applicable for five years. It opened an anti-dumping investigation into bolted and boltless steel shelving from China after finding sufficient prima facie evidence of dumping and injury, and instructed customs authorities to register imports. In mid-September it ordered registration of welded steel mesh imports from China and Turkey pending an anti-dumping investigation, a step that preserves the option of retroactive duties.
Alongside that, the Commission has been extending the geographic reach of European trade policy in the opposite direction. On September 11, 2026 it presented proposals to the Council for the signature and conclusion of the free trade agreement between the European Union and India, an agreement it describes as the largest ever concluded by either party, and which would eliminate or reduce tariffs on 96 percent of European goods exports to India. On the same day it presented the European Union and Ecuador Sustainable Investment Facilitation Agreement to the Council. On September 9 it welcomed Cambodia’s accession to the Multi-party Interim Appeal Arbitration Arrangement, the workaround that a subset of World Trade Organization members use to obtain binding appellate review while the Appellate Body remains without a quorum.
The picture that emerges is of an institution simultaneously liberalising and defending, opening preferential access for partners while raising the cost of entry for goods it considers unfairly traded. Those two tracks are not contradictory in policy terms, but they collide at the level of the individual shipment, and reconciling them is the work of the importer.
Implications for supply chains
Three operational conclusions follow for companies that move goods into or through the European Union.
Duty risk must be modelled as a continuous variable rather than a discrete event. With 232 measures in force, 32 new investigations a year and a surveillance function scanning import data in real time, the probability that any given intermediate industrial good attracts a measure within a five-year sourcing horizon is no longer negligible. Sourcing decisions made on landed cost alone, without a trade defence risk premium, will be wrong often enough to matter.
Origin documentation is now a core compliance discipline rather than a clerical one. The melt and pour evidence rules for steel are the sharpest current example, but the logic is spreading. Where a measure attaches to a material rather than a tariff line, the importer must be able to prove the material’s provenance through every stage of processing.
Diversification away from a single origin is a hedge against trade defence risk as well as against logistics risk. The concentration figures cited in the electrical steel case, where China supplied more than half of European Union imports of the product and its downstream derivatives in 2025, are precisely the profile that attracts measures. Buyers who have already qualified alternative suppliers absorb a measure as a price change. Buyers who have not absorb it as a supply interruption.
Reading the report as a forecasting tool
Trade defence reports are usually filed and forgotten. This one repays treatment as a forecasting document, because the structure of European trade defence makes future action reasonably predictable from present data.
Three variables drive initiation. The first is import penetration, meaning the share of European consumption supplied from outside the bloc, and the rate at which that share is changing. The second is price undercutting, meaning the gap between the landed price of imports and the price at which European producers are selling. The third is the financial condition of the European industry, measured through capacity utilisation, profitability and employment. Where all three deteriorate together in a sector with an organised industry association, a case follows within roughly twelve to eighteen months.
Every one of those variables is now observable to the Commission in near real time through the surveillance task force, and most of them are observable to companies as well through published trade statistics and company reporting. An exporter that monitors its own European market share against European producers’ published capacity utilisation has most of the information needed to anticipate a complaint.
The report also gives a sense of conversion rates. Thirty-two initiations and twenty-six definitive measures in the same year are not the same cohort, since cases take twelve to fifteen months to complete, but the ratio over time is instructive. European trade defence investigations, once initiated, result in measures in the large majority of cases. Termination without measures happens, usually where the complaint is withdrawn or where injury cannot be established, but it is the exception. For planning purposes, an initiation should be treated as a probable measure rather than as an open question.
Duration should be treated the same way. Measures nominally run for five years, but expiry reviews are initiated in most significant cases and, once initiated, keep the duties in force during the review. Several of the 232 measures in force at the end of 2025 have been in place, through successive reviews, for fifteen years or more. The realistic planning horizon for a measure on an intermediate industrial good is a decade, not five years.
Finally, the report’s treatment of enforcement points to where the growth in workload is going. Anti-circumvention investigations, absorption investigations and registration of imports pending investigation are all tools that extend existing measures rather than creating new ones, and all three have been used more frequently in recent years. The registration of welded steel mesh imports from China and Turkey in mid-September 2026, taken while an anti-dumping investigation was still running, is an example of the pattern. Registration is inexpensive for the Commission and creates substantial risk for the importer, since it preserves the possibility of collecting duties retroactively on goods that have already been cleared and sold.
What to watch
The report itself carries no legal force. It is an account, submitted to the European Parliament and the Council, of what the Commission did. Its value lies in what it signals.
Three signals stand out. The first is the sectoral emphasis: chemicals and steel remain the twin engines of European trade defence, and exporters in those sectors should plan accordingly. The second is the institutionalisation of surveillance, which shortens the warning period between a surge and a measure to something close to a single quota season. The third is the absence of provisional anti-subsidy measures in 2025, which sits uneasily against the political salience of subsidy complaints and which will be worth watching in the 2026 figures.
The Commission publishes the full report as document COM(2026)475, with a staff working document numbered SWD(2026)277. For any company with European exposure in a contested sector, the annexes listing measures by product and origin are the single most useful pages of trade policy documentation produced in Brussels each year.