What the EU Machinery Regulation actually changes, and why “digital instructions for use” isn’t what most people think it is.

On 20 January 2027, Regulation (EU) 2023/1230 replaces the Machinery Directive 2006/42/EC. If you’ve seen 14 January 2027 referenced elsewhere: the original Official Journal text contained a date error, which was corrected by a corrigendum (OJ L 169/35, 4 July 2023). The date that applies is 20 January 2027.
Discussion of the Regulation focuses almost entirely on one relief measure: instructions for use may now be provided digitally. This is mostly read as a cost reduction. Less printing, less paper, less logistics.
For companies with a single-language product portfolio, that may be true. For everyone else, the opposite is the case.
First, what is not new, because this is where most misconceptions arise:
The language requirement itself is not being tightened. Art. 10(7) requires that the instructions for use, the safety information, and the information under Annex III be drawn up “in a language which can be easily understood by users, as determined by the Member State concerned.” The same applies to the assembly instructions for partly completed machinery (Art. 11(7)), and to the EU declaration of conformity (Art. 21(2)). There is no obligation to translate into every official EU language. What matters is solely the individual Member State in which the machine is placed on the market, made available, or put into service. In principle, this decentralized mechanism matches existing practice.
So if you were hoping, or fearing, that the Regulation would change your language list: it doesn’t.
What actually does change, and what generates documentation volume:
Availability obligation for the lifetime of the machine, but at least 10 years from placing on the market.
Only for machinery intended for non-professional users, or that can also be used by them.
On request at the time of purchase, free of charge, within one month.
New content subject to documentation requirements that simply didn’t exist before.
Mandatory third-party assessment, significantly expanded technical documentation.
Whoever carries out a substantial modification is considered the manufacturer and bears the full documentation obligation.
Translation demand isn’t rising because new languages are required. It’s rising because new source text is being created, and because virtually every portfolio has to be reviewed and reissued once before the deadline.
19 July 2023
Regulation (EU) 2023/1230 enters into force.
20 January 2024
The provisions on notified bodies already apply. Conformity procedures can be brought forward.
20 October 2026
Member States notify the Commission of their penalty rules. From this point, you know exactly where you stand in your target markets.
20 January 2027
Date of application. The Machinery Directive 2006/42/EC is repealed.
after that: at least 10 years
Every digitally provided set of instructions for use remains accessible, current, and consistent across every market language. This period runs per machine, not per portfolio.
That’s one half of the story. The more interesting half lies elsewhere.
A printed set of instructions for use is a shipment. It’s created, printed, included with the machine, and the job is done. An error from 2019 sits in a binder at the customer’s site. Unpleasant, but static.
A digital set of instructions for use is not a shipment. It’s a service you operate for ten years.
Art. 10(7)(c) requires the digital instructions to remain accessible online “during the expected lifetime of the machinery or related product and for at least 10 years after the placing on the market.” Point (b) additionally requires that users be able to download, save, and print them, and that they remain accessible even during a breakdown of the machine. The same period applies to the EU declaration of conformity when provided digitally (Art. 10(8)).
This shifts something fundamental. Documentation used to be a project with an end date. It becomes an inventory with a status.
And an inventory with a status has a property a shipment doesn’t have: it can drift out of sync.
Run the numbers, not as a marketing figure, but as an operational reality.
You run 40 machine types. You supply 14 markets. That’s 560 valid language versions that must remain accessible, accurate, and mutually consistent for ten years.
Now a safety-relevant passage changes. A warning is made more precise, a limit value is corrected, a residual risk is added. In the German source, that’s an hour’s work.
That’s when the real process starts: 13 language versions have to be updated, reviewed, approved, reformatted, and republished, in such a way that afterward no version with the old status is still accessible.
In the paper world, version drift was a latent risk. In the digital world, it’s a published state. You host the discrepancy yourself.
This is exactly the point most transition projects overlook: the approval loop, not the translation, is the bottleneck. Translation can be done quickly today. Reviewing, approving in layout, and publishing simultaneously in 14 languages: that’s the process that, in most organizations, was never built for repetition, because until now it never had to be repeated.
Until now, a missing or deficient language version was a finding that required an on-site inspection. Someone had to see the machine, open the binder, check the enclosure.
Once the instructions are online, that hurdle disappears entirely. A market surveillance authority, a competitor, or a dissatisfied customer can check the state of your multilingual documentation with a browser. At any time. Without notice. At no cost.
Some caution is warranted in assessing the consequences here, since the Regulation does not explicitly name language violations as a distinct offense. Art. 46 (“Formal non-compliance”) is, however, likely to cover such cases via point (g), which refers to other administrative requirements under Art. 10. If the non-compliance is not remedied, Member States are required under Art. 46(3) to restrict or prohibit the product’s availability, or to arrange for its recall or withdrawal. The penalty framework itself is set by the Member States; it must be “effective, proportionate and dissuasive” and may include criminal penalties for serious infringements (Art. 50).
A final legal assessment of your specific case still requires legal review. For planning purposes, a simpler observation is enough: the cost of an inconsistent language version is rising, while the cost of discovering it is approaching zero.
Four consequences, regardless of which tools you work with:
As long as translation is managed as an assignment with a start and an end, no one is responsible for what happens afterward. With a ten-year availability obligation, that period is exactly the one that matters. Clarify who is responsible for the ongoing state of your 560 versions.
A publication cycle that runs twice a year inevitably creates windows in which the German version is correct and the Polish version is outdated. When both are online at the same time, that window is no longer an internal problem.
Not every change has to trigger every language. But safety-relevant ones do. Clearly marking this distinction in the source document is the cheapest measure with the biggest leverage. It has to happen before the transition, though, not after.
The paper copy on request is tied to two conditions that are easy to overlook: the request must be made at the time of purchase, and delivery must be free of charge within one month. That’s a requirement for the sales process, not for technical writing. And for machinery that can also be used by non-professional users, the obligation remains to include the essential safety information in paper form.
Experience shows the last question is the most uncomfortable one. It’s also the only one a tool can’t solve.
InTO is our platform for multilingual documentation. It translates documents automatically, including InDesign, Word, PowerPoint, Excel, and XLIFF, and moves review and approval directly into the original layout. This eliminates the step that has so far caused the largest share of the effort: text no longer has to be exported, translated externally, and then manually rebuilt into the layout. Dozens to hundreds of technical documents become finished language versions in a short time this way.
What this means in cost terms is shown by our reference customer Hörmann KG, a manufacturer of doors and gates with documentation in over 30 languages. After introducing InTO, its DTP costs fell by 58%, and total translation costs by 69%. These figures come from a layout-intensive InDesign portfolio that previously relied entirely on external DTP. Starting from a different baseline, the numbers look different: for companies with 10 to 15 languages and partly in-house DTP, we conservatively estimate a 40 to 60% reduction in total costs.
For the requirements this article is about, a second point matters more: the cost of updates.
Instructions for use that are maintained for ten years aren’t translated once, they’re translated again and again, in small increments. A warning refined here, a paragraph added there, a new component in chapter 7. If every one of these changes is treated like a new order, over the years you pay many times over what the actual scope of the change would justify.
This is exactly what InTO is built for. Text passages that have already been translated and approved are reused with every update. Only what has actually changed gets translated and reviewed. The same applies to layout: existing language versions are carried forward and updated, rather than rebuilt from scratch on every run. For a small change across 14 languages, that’s the difference between an afternoon and a project.
Over a ten-year provision period, the decisive cost block is therefore not the first translation, but the sum of every update that follows.
In about an hour, we’ll look together at how your multilingual documentation comes together: which formats are in use, how many language versions you maintain, how often updates happen, and where approval loops cost you time.
Above all, what the process actually costs, including the internal effort that never shows up on any agency invoice. You’ll then receive a written evaluation with a solid cost estimate and an assessment of where you stand with the deadline in view.
We weigh the cost of introducing InTO against your concrete savings potential.
We only recommend an implementation if it produces a solid return on investment.
Request AnalysisSources
This article reflects the legal position at the time of publication and does not constitute legal advice.