Originally published on Medium.
The registry has been live since July. The first hard deadline is 18 February 2027, and the law behind it is late.

From 18 February 2027, certain batteries cannot legally be placed on the European market unless a digital passport for them is registered in an EU database. It has been open since 20 July, and customs can already query it before goods are released.
The batteries are the first. Iron and steel are scheduled for a delegated act in the last quarter of this year, textiles, tyres and aluminium in 2027, furniture in 2028, mattresses and ICT products in 2029, each with at least eighteen months before the obligation bites.
That is the part everyone in a compliance function already knows, and it is not why this is worth your morning. The interesting question is the one the regime quietly depends on and has never answered: when a product arrives at the European border with no passport and no record, who is the authority supposed to call?
Europe’s own numbers say that for a large share of what arrives, the answer is nobody.
What is actually live
The European Commission published a press release on 20 July 2026 under the headline that the Digital Product Passport Registry is now live. Its text is unglamorous: economic operators “must register each Digital Product Passport in the Registry,” and registration “is available through either a secure user interface or an application programming interface (API).”
The registry does not hold your product data. The Commission describes it as storing “unique identifiers, registration data, and high-level metadata,” with the product information itself staying with the operator, who “may host these themselves or through DPP service providers.”
The legal basis is the Ecodesign Regulation, Regulation (EU) 2024/1781. The battery deadline comes from a different instrument, the Batteries Regulation, Regulation (EU) 2023/1542, and covers electric vehicle, light means of transport and certain industrial batteries.
One detail matters more than its length suggests. On the carrier, the Commission says only that “a digital data carrier, such as a QR code, links the physical product to its DPP.” It does not mandate a technology. A printed QR code satisfies it.
The border, and the person standing at it
Picture the officer, because this is where the regime stops being paperwork.
Under the customs arrangements the Commission describes, an officer can verify electronically that an imported product has a valid registered DPP and that the commodity code has been provided, before the goods are released for free circulation. That is a database lookup, not an inspection. It takes no laboratory, no sample and no expertise in batteries.
Which makes it the first enforcement tool in this field that actually scales to the volume. And the volume is the whole problem. Low-value consignments into the EU ran to 1.4 billion in 2022, 2.4 billion in 2023 and 4.6 billion in 2024, according to figures the European Parliament’s research service attributes to the single market strategy. For 2025 the Commission’s own customs reform page gives 5.9 billion, over ninety per cent of it from China. E-commerce now accounts for 97 per cent of all customs declarations.
Against that, physical checking has been going the other way. Customs checks fell to 82 per million imports in 2024, from 124 in 2023 and 203 in 2022. Market surveillance authorities checked 102 products per million inhabitants in 2024. The Commission’s own 2026 single market report calls the level of market surveillance insufficient.
So a lookup that can run on every consignment is worth more than an inspection regime that reaches almost nothing. That is the case for the passport, and it is a good one.
What Europe found when it looked
Three sets of figures, all from the European Parliament’s research service briefing on the European product act, each attributed there to a Commission source.
In a sweep announced on 18 July 2025, the Commission reported that around 52 per cent of checked listings carried compliant manufacturer information, rising to 65 per cent on very large online platforms. On information about the responsible person established in the EU, 68 per cent of listings were compliant, and 81 per cent on the very large platforms. After a product was removed by a very large platform, the briefing records, it reappeared in six cases.
In a priority control area running from April to June 2025, covering toys and small electronics bought from four major online marketplaces, 55 per cent of products were non-compliant overall and 84 per cent of those lab tested were found to be dangerous.
And Safety Gate, the EU’s alert system, logged 4,671 alerts for dangerous non-food products in 2025, up 13 per cent on 2024 and 121 per cent on 2022. Of those, 2,268 concerned products from Asian countries and 2,006 specifically from China.
Read the second figure again, because it is the one people skip. Not 84 per cent of a random sample. Eighty-four per cent of the products that were selected for laboratory testing turned out to be dangerous.
The gap the passport sits on top of
Here is the structural problem, and it is the clearest single fact I found this week.
BEUC, the European consumer organisation, published a checklist in June 2026 setting out what it wants from the forthcoming law. In it, BEUC states that the Market Surveillance Regulation requires an economic operator established in the EU for only 20 of 71 harmonised product categories. For non-harmonised products under the General Product Safety Regulation, the same requirement does not apply at all.
Twenty of seventy-one. For the rest, there is no legal obligation that anybody inside the European Union be answerable for the product.
A digital product passport is a record created and registered by an economic operator. If no economic operator is required to exist, the passport regime has a hole in the middle of it exactly where the parcels are arriving.
BEUC’s ask follows from that: online marketplaces should become economic operators where no manufacturer, importer or authorised representative in the EU exists or can be found.
The law that was due in September
Which brings us to the European Product Act, and to the reason this piece leads with a battery deadline rather than a bill.
The Commission’s 2026 work programme set adoption of the package for the third quarter of this year. The Parliament’s research service noted in June that the likely date was September. Law firms tracking it expected publication on 6 October. Cooley’s own tracker now carries a banner saying publication has been postponed until later this year, date to be confirmed. As of today it has not appeared.
What is known comes from a leaked draft, which is a weaker thing than a bill and should be read that way. Three firms have published analyses of it, and they agree on the shape. The Act would merge the New Legislative Framework and the Market Surveillance Regulation, Regulation (EU) 2019/1020, into one directly applicable regulation. Noerr puts the draft at over 220 pages.
The provision that matters for everything above: a marketplace would have to verify, before a listing goes live, that the product has a valid digital product passport or product responsibility record and that the listing matches it. If it does not, and the product reaches the market with no EU-established operator responsible for it, the marketplace itself assumes the obligations of an authorised representative.
On penalties, the sources differ and the difference is worth noting rather than averaging. Cooley reads the draft as requiring member state penalties of between 1 and 6 per cent of global annual turnover, with separate Union-level penalties of up to 6 per cent. Noerr reports fines of up to 6 per cent of global annual turnover. Nobody is quoting a published text, because there isn’t one.
At a European Parliament committee hearing in early October, a Commission official said the Commission is considering requiring every imported product to have a responsible operator within the EU. That is the gap being closed, said out loud.
What it would have cost with unit-level identity
Here is the honest commercial read, and it has two halves that point in different directions.
The first half is that this is a dated project with a known scope, and most firms in scope are treating it as a labelling exercise. The work is: assign a unique identifier to every unit, hold a record against it, publish a resolvable link, register the identifier. A company that does the minimum produces a QR code that opens a PDF. A company that does the same work once, properly, ends up with the thing that sets its recall width, because an identifier that resolves to a unit is the same asset whether a customs officer queries it or a quality director does at two in the morning. The deadlines are already written down. February for batteries, then a category a year.
The second half is the part our own industry keeps quiet about. The regulation does not require NFC, and it does not require anything we sell. The Commission’s wording is “a digital data carrier, such as a QR code.” A label printer satisfies the law. Anyone telling you the EU is mandating secure authentication is selling you something, and we are not going to be that.
What the passport establishes is that a registered record exists and that this code points at it. It does not establish that this physical item is the one the record describes. A QR code is a number printed on a label, and a number printed on a label can be printed again on another label. We covered the limits of what a scan proves earlier in this series and the regulation does not change them.
So the useful question for a compliance budget is not whether to comply. It is whether the identifier you create to satisfy February is one you can also rely on in a recall, in a customs dispute, or when somebody shows you a unit and asks whether it is yours. If you want to know which of those your current records support, the recall width assessment takes about three minutes and collects nothing.
What we are not saying
We are not saying the European Product Act will pass as drafted, or at all. It is an unpublished leaked draft, the proposal is already late, and five member states, the Netherlands, Finland, France, Portugal and Sweden, put a joint note to the Competitiveness Council on 22 May 2026 asking among other things that physical CE marking be kept, which is not the direction the draft goes. Legislative drafts change, and this one has not even been tabled.
We are not saying a digital product passport reduces counterfeiting. It makes the absence of a record detectable at a border, cheaply and at volume, which is a different and narrower thing.
And we are not saying the figures above are a measurement of counterfeit goods. They measure non-compliance and danger, which overlap with counterfeiting without being it. A product can be perfectly genuine, made by exactly who it says, and still fail a toy safety standard.
Aeroz tracks counterfeit enforcement, recalls and supply chain failures in regulated industries, three times a week. Follow for the next one. aeroz.io · [email protected]
