Digital Operational Compliance & EU AI Act Knowledge Base

Product frameworks for DORA, NIS2, EU AI Act, operational resilience, compliance controls, audit evidence and digital risk workflows.

The 10^25 FLOP threshold: what the AI Act never says

Article 51(2) of the EU AI Act (Regulation (EU) 2024/1689) presumes a general-purpose AI model has high-impact capabilities once the cumulative compute used to train it, measured in floating-point operations, exceeds 10^25. What the Act never spells out is how to actually count that number. Do earlier versions of a model contribute their compute to a later one’s tally? If a model is branched, is each branch’s compute counted separately or added together? A threshold this consequential, sitting with no counting methodology attached to it, is the whole story here.

What the threshold is and isn’t

Article 51(1) classifies a general-purpose AI model as presenting systemic risk if it meets either of two conditions: it has high-impact capabilities established through appropriate technical tools and methods, or the Commission has designated it as comparable, ex officio or following a qualified alert from the scientific panel, based on the criteria in Annex XIII. The 10^25 FLOP figure in Article 51(2) only speaks to the first route — it’s a rebuttable presumption that high-impact capabilities exist, not a definition of systemic risk itself, and not the only way a model ends up in scope. Article 51(3) lets the Commission adjust the threshold by delegated act to reflect the state of the art, including algorithmic improvements and more efficient hardware — so even the number itself isn’t fixed forever, only fixed until the Commission decides otherwise.

The contradiction at the centre

Recital 111 defines high-impact capabilities relative to the frontier: capabilities that match or exceed those recorded in the most advanced general-purpose AI models. That’s a moving target by design — as the frontier advances, the bar for “high-impact” moves with it. The 10^25 figure Article 51(2) actually operationalises this with is nothing of the sort: it’s fixed, absolute, and doesn’t move on its own between one delegated act and the next.

The practical consequence is exactly the one the Act’s own drafting implies but never resolves: a model that clears the relative, frontier-based bar this month may not clear it next month once competitors release something more capable, while the absolute FLOP number attached to that same model never changes at all. There’s a second problem sitting underneath this one — most frontier models are proprietary, so the actual comparator Recital 111’s test asks you to measure yourself against usually isn’t measurable at all. The relative definition and the absolute number were clearly meant to approximate each other. Nothing in the text guarantees they track.

The counting questions the text leaves open

Three practical questions decide whether a lot of real models sit above or below the line, and the Act answers only one of them.

Fine-tuning — this one has a real answer. Recital 111 states that the cumulative compute input used for training includes the total compute used for activities to improve the model’s capabilities prior to deployment, explicitly naming pre-training, synthetic data generation, and fine-tuning. So fine-tuning compute counts toward the cumulative total, at least for the fine-tuning that happens before a model’s initial deployment. What that doesn’t resolve is the harder version of the same question: whether fine-tuning or continued training applied after an earlier version was already placed on the market gets added to that version’s original tally, or starts a new count.

Prior iterations — genuinely unresolved. Nothing in the text says whether compute spent on earlier versions of a model contributes to a later version’s cumulative figure, which matters enormously for any provider that iterates on the same underlying architecture across releases rather than training each version from nothing.

Branches — also unresolved. If a single training run forks into multiple variants, nothing says whether each branch’s compute is counted independently or whether the branches’ compute is summed as one cumulative figure against the threshold.

Annex XIII gives the Commission its own set of proxies for the designation route — estimated training cost, time, or energy consumption, as alternatives to a direct FLOP count — but it’s unclear whether a provider doing its own self-assessment under Article 51(2) has to use the same criteria, and none of these proxies answer the iteration or branching questions regardless of which measure you use.

The other Annex XIII numbers

The FLOP threshold gets all the attention, but Annex XIII gives the Commission six other criteria to weigh when designating a model as comparable under Article 51(1)(b), and one of them is rarely quoted at all:

  • the number of parameters in the model;
  • the quality or size of the training dataset, for example measured in tokens;
  • the amount of compute used to train it, measured in FLOPs, or expressed through a combination of other variables such as estimated cost, time, or energy consumption;
  • the model’s input and output modalities — text-to-text, text-to-image, multimodal — along with advanced thresholds for high-impact capabilities specific to each modality, and the particular type of input or output involved, such as biological sequences;
  • benchmarks and capability evaluations, including the number of tasks the model performs without additional training, its adaptability to learn new distinct tasks, its degree of autonomy and scalability, and the tools it can access;
  • whether it has a large impact on the internal market because of its reach — which is presumed once the model has been made available to at least 10,000 registered business users established in the EU;
  • the number of registered end-users.

What you do when you cross it

If a model meets the Article 51(1)(a) condition, the provider notifies the Commission without delay, and in any event within two weeks of the condition being met or of it becoming known that it will be met. The notification has to include the information needed to demonstrate the condition is satisfied. Providers aren’t limited to reporting the fact, either — in that same notification, a provider can put forward sufficiently substantiated arguments that, despite meeting the threshold, the model exceptionally doesn’t present systemic risks because of its specific characteristics, and therefore shouldn’t be classified as a systemic-risk model. If the Commission decides those arguments aren’t sufficiently substantiated, it rejects them, and the model is treated as a systemic-risk model regardless.

The appeal loop

The self-notification route comes with a built-in argument mechanism. The Commission’s own ex officio designation route, under Article 51(1)(b), doesn’t have an equivalent express provision letting a provider make its case before the decision is made — which pushes toward providers over-notifying rather than risk being designated without having had the chance to argue first.

Once a model has been designated this way, though, there is a path back. A provider can submit a reasoned request for reassessment, and the Commission will consider whether the model, measured against the Annex XIII criteria, is still properly considered a systemic-risk model. That request has to rest on objective, detailed, and new reasons that came to light after the designation decision — not a restatement of the original case — and the earliest a provider can make it is six months after the designation. If the Commission reassesses and decides to maintain the designation anyway, the provider has to wait another six months before trying again. There’s no cap on how many times this can repeat, only the six-month cooldown between each attempt.

Frequently asked questions

Is 10^25 a hard line?

No, in either direction. Above it, a provider can still argue its way out under Article 51(2) and 52(2)-(3) if it can substantiate that the model’s specific characteristics mean it doesn’t actually present systemic risk. Below it, the Commission can still designate a model under Article 51(1)(b) if it finds comparable capabilities or impact using the Annex XIII criteria. The number sets a presumption, not a verdict.

Who publishes the list of systemic-risk models?

The Commission does. Article 52(6) requires it to publish and keep an up-to-date list of general-purpose AI models with systemic risk, though that duty is itself qualified by the need to respect intellectual property rights and protect confidential business information and trade secrets under EU and national law.

Does the Omnibus change Articles 51-55?

No. This is one of the more consistently confirmed points across current commentary: the Digital Omnibus reform is squarely focused on the Annex III and Annex I high-risk timeline, registration, and related provisions. The general-purpose AI rules in Articles 51 through 55 have applied since 2 August 2025 and remain untouched by the reform.

What if we don’t know our own FLOP count?

The Act doesn’t offer a substitute self-certification route for this situation — the cost, time, and energy proxies in Annex XIII exist for the Commission’s own comparable-capabilities designation under Article 51(1)(b), not explicitly as an alternative counting method a provider can rely on for its own Article 51(2) self-assessment. A provider without confident compute accounting is sitting in exactly the gap this article describes, without a clearly signposted way out of it.

Can we be designated without ever notifying?

Yes. Article 52(1) states plainly that if the Commission becomes aware of a general-purpose AI model presenting systemic risks that it hasn’t been notified about, it can decide to designate that model as one anyway. Notification is the expected route in; it isn’t the only route the Commission can use to get there.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base, RegTech Glossary & Standards

Significant change vs substantial modification: same thing?

Article 111(2) of the EU AI Act (Regulation (EU) 2024/1689) grandfathers high-risk systems already on the market unless they undergo “significant changes” in design after the general application date. Article 3(23), in a completely separate part of the Regulation, defines “substantial modification.” Recital 177 says the two are equivalent in substance. White & Case’s own Handbook commentary on this point is unusually blunt for a law-firm guide: “It is not entirely clear why two separate terms are used.” Whether your legacy AI estate stays outside the Act’s high-risk rules turns on a distinction its own drafters apparently didn’t mean to create.

What the grandfathering actually gives you

Article 111(2) applies the Regulation to operators of legacy high-risk systems — those placed on the market or put into service before the general application date — only if those systems undergo significant changes in their design from that date onward. Read that structure carefully: it isn’t a grace period that eventually expires on its own. It’s an indefinite exemption, conditional on one thing only — that you don’t significantly change the system. Stand still, and the exemption holds indefinitely. Change it, and you’re pulled into the full regime from that point.

The three exceptions to the free pass

Three categories don’t get to rely on standing still.

  • Public authority deployers. Regardless of whether a system undergoes any change at all, providers and deployers of high-risk AI systems intended for use by public authorities have to comply with the Regulation’s requirements no later than six years after entry into force — 2 August 2030.
  • Annex X large-scale IT systems. AI components of the large-scale IT systems established under the legal acts listed in Annex X have their own fixed deadline: 31 December 2030, for systems placed on the market before the relevant cutoff.
  • The Article 5 prohibitions. These were never grandfathered in the first place. Both of Article 111’s grandfathering provisions open with “without prejudice to the application of Article 5” — the prohibited-practices regime has applied since 2 February 2025, full stop, regardless of when any given system was placed on the market.

The GPAI cliff edge

Article 111(3) draws its own hard line, and it’s worth seeing exactly how sharp it is. A general-purpose AI model placed on the market before 2 August 2025 gets until 2 August 2027 to come into compliance. A model placed on the market on or after 2 August 2025 faces enforcement immediately, from the moment it’s placed on the market. White & Case’s own framing of this makes the stakes concrete: a model placed on the EU market on 1 August 2025 faces no enforcement risk for two years; the same model placed one day later, on 2 August 2025, faces immediate risk. Providers of GPAI models were, in the Handbook’s words, incentivised to place those models on the EU market before that specific deadline. One day either side of a date is worth two years.

Does the Omnibus reset the clock?

Article 111(2)’s grandfather clause attaches to “the general date of application” — and the Digital Omnibus reform moved that date, deferring Annex III high-risk obligations to 2 December 2027 and Annex I obligations to 2 August 2028. That’s not a cosmetic shift. If the cutoff itself moves, the population of systems that count as “legacy” moves with it: a system placed on the market in, say, October 2027 — which would have squarely missed the old 2 August 2026 cutoff and faced the full regime immediately — can now fall on the grandfathered side of a later line, provided it isn’t significantly changed afterward.

The honest caveat, consistent with everything else tied to this reform: the new dates bind only once the amending regulation is published in the Official Journal and enters into force. Until then, the original 2 August 2026 cutoff remains the legally operative one. The GPAI timeline in Article 111(3) is a separate matter entirely — nothing in the Omnibus reform touched the 2 August 2025 / 2 August 2027 dates, since that deferral was built specifically around the Annex I and Annex III high-risk regime, not the general-purpose AI rules.

So what is a significant change?

Since Recital 177 treats the two concepts as equivalent, the operative test is the one Article 3(23) actually writes down: a change that was not foreseen or planned in the initial conformity assessment, and that either affects the system’s compliance with the Chapter III, Section 2 requirements, or modifies the intended purpose the system was assessed for. That’s the same test this piece’s companion article on Article 43(4) walks through in detail for systems that keep learning after deployment — the pre-determined change envelope you document there is, in substance, the same boundary that decides whether a legacy system has quietly lost its grandfathering here.

Frequently asked questions

Does a UI change count as a significant change?

Usually not, unless it touches how the system actually meets the Chapter III Section 2 requirements — a change to how oversight information or output explanations are presented to a human reviewer could cross that line, where a purely cosmetic interface refresh wouldn’t. The test is whether compliance or intended purpose is affected, not how visible the change is to an end user.

Does deploying into a new customer segment count?

Likely yes, if the new segment amounts to a different intended purpose than what the system was originally assessed for. This is close to the paradigm case the “modifies the intended purpose” limb of Article 3(23) is written for — the same system, doing a materially different job for a materially different population, is exactly the scenario grandfathering isn’t meant to protect.

Does a routine security patch count?

Generally not, if it’s the kind of maintenance foreseeable at the time of the initial assessment and doesn’t itself change how the system meets the substantive requirements or what it’s for. A patch significant enough to materially alter the system’s behaviour or outputs is a different matter, and the size of the change matters more than its label as “just a patch.”

What if we never carried out a conformity assessment in the first place?

This is a harder problem than a grandfathering question, and this piece won’t pretend otherwise. Article 3(23)’s test explicitly measures a change against what was foreseen or planned in the initial conformity assessment — without one, there’s no documented baseline to test a change against at all. A system in that position has a more fundamental compliance gap than whether a later change was “significant”; that gap is worth resolving before the grandfathering question becomes the relevant one.

Can we deliberately freeze a system’s design to stay exempt?

Yes, and White & Case’s own commentary treats this as a real, foreseeable incentive rather than a loophole — not changing a legacy system’s design keeps it inside Article 111(2)’s exemption indefinitely, subject to the three exceptions above. Freezing the system’s design doesn’t freeze every other obligation that might apply regardless of risk tier, though — Article 4 AI literacy, for instance, isn’t tied to this grandfathering at all and applies whether or not the underlying system changes.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base

Retraining a high-risk model without a new assessment

Article 43(4) of the EU AI Act (Regulation (EU) 2024/1689) provides that, for high-risk systems that keep learning after being placed on the market or put into service, changes to the system and its performance that were pre-determined by the provider at the time of the initial conformity assessment — and that form part of the technical documentation required under Annex IV, point 2(f) — do not count as a substantial modification. No new conformity assessment follows. The entire cost of staying compliant while you keep retraining rests on documentation you write once, at the start, and almost nothing written for MLOps teams says so plainly.

The rule and the timing that makes or breaks it

Read the timing again, because it’s the whole mechanism: the changes have to be pre-determined at the time of the initial conformity assessment. Not documented eventually. Not written up after a retraining cycle already happened and worked out fine. If it isn’t in the technical file the assessment was actually based on, it doesn’t get the exemption — it’s just an undocumented change, and undocumented changes are exactly what Article 3(23) is built to catch.

What “pre-determined” has to look like in a technical file

Annex IV, point 2(f) — the specific provision Article 43(4) cross-references — requires, where applicable, a detailed description of pre-determined changes to the AI system and its performance, together with all relevant information about the technical solutions put in place to ensure the system’s continued compliance with the Chapter III, Section 2 requirements as those changes happen. That’s two things, not one: a description of what can change, and an explanation of how you’ve engineered continued compliance despite it changing.

Two companion provisions in the same Annex turn this from a paperwork exercise into something that has to actually be tested. Point 2(g) requires dated, signed test logs and reports specifically covering the pre-determined changes described in point (f) — you don’t just describe the envelope, you have to show testing that happened inside it. And point 6, separately, requires an ongoing description of the relevant changes the provider actually makes over the system’s lifetime — a running log distinct from the upfront envelope, recording what happened against what was planned.

In practice, the envelope worth drafting into that technical file covers: the retraining cadence itself; the bounds of the data sources involved, not just their identity; the performance ranges you expect, tied to the specific metrics you’re measuring against; drift thresholds that would trigger a stop; rollback conditions if a retrained version underperforms; and, most importantly, a clear statement of the point at which a change would leave this envelope and stop being “pre-determined” at all. Tie each of these back to your Article 11 technical documentation as a whole and your Article 17 quality management system — the envelope isn’t a standalone document, it’s a component of both.

What is definitely outside the envelope

Article 3(23) defines a substantial modification as a change to an AI system, after it’s placed on the market or put into service, that the provider did not foresee or plan in the initial conformity assessment, and that either affects the system’s compliance with the Chapter III, Section 2 requirements, or changes the intended purpose the system was assessed for. Two disjunctive triggers, either one sufficient, sitting on top of the same threshold question: was this foreseen at the time of the initial assessment or wasn’t it.

A retraining run using the data sources, cadence and performance bounds you described in your point 2(f) envelope, producing a model that still does the same job to the same standard, sits inside the exemption — that’s precisely what Article 43(4) exists to cover. A retraining run that pulls in a materially different data source you never documented, or that shifts the system toward a new use case, sits outside it on both available grounds: it wasn’t foreseen, and it plausibly touches compliance or purpose. The difference isn’t how big the change feels. It’s whether you wrote it down before you were assessed.

Where the envelope has to be re-declared

Once a change actually crosses into substantial modification, the practical consequence is that a new conformity assessment applies — the system is, in effect, being placed on the market again in its modified form. That loops back to registration: the EU database registration duty attaches to placing a system on the market or putting it into service in the first place, so a substantially modified system re-triggers that same duty rather than quietly inheriting the original entry. Treat a substantial modification as resetting the compliance clock, not just the model weights.

The neighbouring clocks

A few retention and monitoring duties sit right next to this one, and they’re worth having straight rather than confusing with each other. Article 12 requires high-risk systems to automatically log events across their lifecycle, and one of the three purposes that logging capability is explicitly built for is identifying situations where a substantial modification occurs — the log isn’t just an audit trail, it’s part of how you’re meant to notice you’ve left your own envelope. Article 19 then sets how long those logs have to be kept: a period appropriate to the system’s intended purpose, and at least six months, unless other Union or national law — particularly data protection law — requires longer. Separately, Article 18 requires the provider to keep the technical documentation, the quality management system documentation, any documentation of changes approved by a notified body, notified-body decisions, and the declaration of conformity available to national authorities for ten years after the system was placed on the market or put into service.

Frequently asked questions

Does a fine-tune count as a substantial modification?

It depends entirely on whether it falls inside your documented envelope. A fine-tune using the data sources and staying within the performance bounds you described in your point 2(f) file isn’t a substantial modification — that’s the exemption working as intended. A fine-tune that pulls in undocumented data or pushes performance outside the ranges you committed to isn’t covered, regardless of how routine it felt to run.

Does swapping the base model count?

Almost certainly, yes. Swapping the underlying model is rarely something a provider pre-determined as a continued-learning change in the original technical file — it’s closer to a different system than a retrained version of the same one, and it’s hard to argue it doesn’t touch Chapter III Section 2 compliance or the assessed intended purpose. Treat a base-model swap as a substantial modification by default, and build the case for an exception only if your original documentation genuinely anticipated it.

Does a new training run on the same pipeline count?

Only if it stays inside what you documented. The same pipeline producing a run that matches the data-source bounds and performance ranges already on file is exactly what the exemption is for. The same pipeline producing a run against new data, or aimed at different performance characteristics than what you described, has left the envelope even though nothing about the infrastructure changed.

Who decides — us or the notified body?

For Annex VI self-assessed systems, the provider makes this call itself, as part of its ongoing Article 17 quality management obligations — there’s no notified body in the loop to consult. Where a notified body is involved under Annex VII, changes it has specifically approved get their own retention requirement under Article 18, which suggests an ongoing role in confirming certain categories of change — but the initial pre-determination is still the provider’s own documentation choice, made before that body ever assessed anything.

Does the December 2027 date change any of this?

Not the substance. Article 43(4) is part of the Chapter III framework that, for Annex III systems, now applies from 2 December 2027 rather than 2 August 2026 following the Digital Omnibus reform. The rule itself is unchanged — only the date it starts binding on stand-alone Annex III systems moved, alongside every other obligation in that same package.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base

You still register the AI system you decided isn’t high-risk

Under Article 6(4) of the EU AI Act (Regulation (EU) 2024/1689), a provider who concludes under Article 6(3) that its Annex III system is not, in fact, high-risk still has to register that system in the EU database — with a short summary of the grounds on which it reached that conclusion. That entry sits in the public section. The Commission’s Digital Omnibus proposal of 19 November 2025 tried to delete this obligation outright. Both the Council, in its negotiating mandate of 13 March 2026, and the European Parliament’s IMCO and LIBE committees, in a joint report adopted five days later on a 101-9-8 vote, independently rejected the deletion. The final text reinstates registration, with streamlined content requirements. If your compliance plan was drafted off the November proposal, it’s currently wrong.

The rule in one paragraph

Exemption doesn’t mean invisibility. Concluding that your Annex III system clears the Article 6(3) bar doesn’t end your paperwork — it starts a different kind. You keep the internal assessment that got you to that conclusion, you file a summary of it in the EU database, and that summary sits where the public can read it.

The three database categories

The EU database sorts entries into three categories, and who files depends on which one applies:

  • Category 1 — genuinely high-risk systems under Article 6(2) and Annex III. Filed under Annex VIII Section A. The provider files.
  • Category 2 — systems a provider has assessed as not-high-risk under Article 6(3). Filed under Annex VIII Section B. The provider files.
  • Category 3 — high-risk systems used by public authority deployers. Filed under Annex VIII Section C, including a summary of the data protection impact assessment carried out under Article 35 GDPR. The deployer files, not the provider.

Category 2 is the one that catches people off guard, precisely because “not high-risk” sounds like an off-ramp from the database entirely. It isn’t.

What your competitors get to read

Category 2 entries go into the public section of the database, the same section genuinely high-risk registrations sit in. The public section is free to access, navigable, and machine-readable. That’s not a side effect — it’s the design.

The practical consequence is straightforward and easy to underestimate: the legal argument you’re relying on to stay out of the high-risk regime is a document a competitor, a journalist, or an NGO can read, and can challenge. If your reasoning under Article 6(3) is thin, it’s thin in public.

The Article 80 backstop

Article 80 gives market surveillance authorities a specific procedure for exactly this situation, and it’s worth reading in full rather than taking on faith. Where an authority has sufficient reason to believe a system a provider classified as not-high-risk is actually high-risk, it tests that classification against the Article 6(3) conditions and the Commission’s guidelines. If the test confirms the system is high-risk, the authority orders the provider to bring it into compliance within a deadline the authority sets. Miss that deadline, and fines follow under Article 99 — that’s the backstop the brief refers to, and it’s a real, specific consequence, not a vague risk of reputational harm.

There’s a second, sharper consequence sitting one paragraph later. If the authority’s review finds the provider classified the system as not-high-risk specifically to circumvent the Chapter III Section 2 requirements — not a good-faith misjudgment, but an evasive one — that draws its own, separately finable violation under Article 99.

And Article 80 names its own evidence source. In exercising their oversight, market surveillance authorities may carry out checks that take into account, in particular, information stored in the EU database. That’s not a general observation about transparency — it’s the Act telling authorities where to look first. Your Category 2 summary isn’t a filing that disappears into an archive; it’s a named input into exactly the kind of review Article 80 describes.

What the Omnibus changed

The sequence is worth having straight, because it moved more than once. The Commission’s Digital Omnibus package, published 19 November 2025, proposed deleting the Article 6(4) registration obligation for self-assessed not-high-risk systems entirely. The Council’s negotiating mandate of 13 March 2026 rejected that specific deletion while broadly aligning with the Commission elsewhere, reinstating a simplified registration obligation. Five days later, on 18 March 2026, the Parliament’s IMCO and LIBE committees adopted a joint report doing the same, by a 101-9-8 vote. A provisional political agreement reached on 7 May 2026 confirmed the reinstatement, with streamlined Annex VIII Section B content — fewer data points required in the filing, without removing the filing itself. Parliament adopted the final text on 16 June 2026, and the Council gave its final green light on 29 June 2026.

What’s actually different in the streamlined Section B, line by line, isn’t something I can confirm precisely — commentary consistently describes it as a reduced set of required data points rather than a restructured form, but I haven’t seen the specific before-and-after list. What isn’t in question is the outcome: the obligation survived a deletion attempt from its own drafter, unanimously rejected by both co-legislators before trilogue even started.

One more date worth anchoring this to: Annex III high-risk obligations, under the same reform, now apply from 2 December 2027 rather than the original 2 August 2026. The registration duty for Category 2 systems tracks that same timeline.

The one Annex III category that escapes the EU database

There’s a genuine exception, and it’s narrow. High-risk AI systems falling under Annex III point 2 — critical infrastructure — register at national level instead of in the EU database. The Act doesn’t name a single EU-wide national register to check; the obligation is simply pushed down to whichever Member State mechanism applies, and that mechanism isn’t uniform across the Union. If your system is critical-infrastructure-adjacent, “check the EU database” isn’t the right instruction to give your compliance team.

Frequently asked questions

Do I register before or after placing the system on the market?

Before. Registration is structured as a precondition tied to placing the system on the market or putting it into service, not a filing you make afterward to tidy up the paperwork. Treat the Article 6(3) assessment and the database entry as part of your go-live checklist, not a follow-up task.

What if my system triggers the profiling override?

Then Article 6(3) isn’t available to you at all. Any Annex III system also used to profile natural persons is high-risk regardless of how narrow, preparatory, or human-reviewed its role looks otherwise — there’s no exemption to claim. You’re in Category 1, filing under Annex VIII Section A, not Category 2.

Can I redact the grounds I file?

Not for a standard Category 2 entry — the summary of your grounds is what goes in the public section, and that’s the point of the filing. Restricted, non-public treatment is reserved for specific categories named in the Act, chiefly biometrics, law enforcement, migration and border control, and similar sensitive uses. A conventional HR, credit, or education system claiming the Article 6(3) exemption doesn’t get that treatment.

Who registers if we’re the importer, not the original developer?

Ordinarily, the original provider — including one established outside the EU, acting through its EU authorised representative. You become the provider yourself, with the provider’s registration duty, only if you put your own name or trademark on the system, make a substantial modification to it, or modify a non-high-risk system’s intended purpose in a way that makes it high-risk. A straightforward import without any of that doesn’t shift the filing obligation onto you.

Does this apply to a system already on the market?

This is genuinely less settled than the rest of this piece. The Act’s general grandfathering rule protects high-risk systems already on the market from the new obligations unless they undergo significant changes, and the same logic would plausibly extend to a system you’d assessed, before the relevant date, as not needing Article 6(3) registration at all. But I haven’t found a source that addresses this specific edge case directly, so treat it as a reasonable inference rather than a confirmed answer, and check it against the Commission’s guidance before relying on it.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base

The AI Act reaches companies that never targeted the EU

Recital 22 of the EU AI Act (Regulation (EU) 2024/1689) frames the Act’s reach beyond the EEA around output “intended to be used” there. Article 2(1)(c), the operative provision that actually does the work, drops the word “intended” entirely: it applies to providers and deployers established outside the EEA “where the output produced by the system is used in the [EEA].” White & Case’s own Handbook calls the two “inconsistent,” and works through what that inconsistency actually does to a business that never meant to come near EU law. The argument worth making explicitly, rather than leaving implicit: the operative text creates a jurisdictional trigger with no off-switch built in, and the recital that would have installed one has no binding force of its own.

The claim

Read literally, Article 2(1)(c)’s territorial trigger is an event downstream of the operator’s own control — a later, independent decision by someone else about where to use an AI system’s output — rather than a choice the operator itself makes. That is a genuinely unusual jurisdictional design for EU law, and it is not what Recital 22 says the rule is supposed to be.

How this differs from the GDPR, precisely

Article 3 GDPR turns on offering goods or services to, or monitoring, individuals in the EEA — both are tests shaped around the entity’s own intent or conduct. You come within the GDPR’s territorial scope by doing something aimed at the EEA. Article 2(1)(c) isn’t built the same way: nothing in its text asks what the provider or deployer meant to do. This is worth stating carefully rather than overstating — it’s a difference in what triggers the rule, not necessarily a difference in how ambitious EU lawmakers were being. The GDPR and the AI Act both clearly intend broad reach; only one of them ties that reach to the regulated party’s own choices.

The worked example, in full

White & Case’s Handbook walks through exactly this scenario. An advertising agency based in Japan uses third-party AI systems in its ordinary workflow to generate branding concepts for clients. A customer based in Argentina commissions branding for one of its products; the agency delivers it, using AI to generate some elements; the client is happy and pays. A year passes. The Argentine client then opens a new office in Spain, and uses that branding there.

Under a literal reading of Article 2(1)(c), the Japanese agency is the deployer of the AI system it used to produce the branding; the branding itself is the “output”; the client has now used that output in the EEA. The agency — based solely in Japan, with no intention of ever doing business in Europe — falls within scope, a year after the work was finished, because of a decision it had no part in and no visibility into. White & Case’s own conclusion is blunt: there does not appear to be any way for the agency to avoid this outcome. A contractual clause prohibiting the client from using the branding in the EEA wouldn’t help, because Article 2(1)(c) doesn’t appear to take intent into account at all — and a restriction aimed at intent has nothing to grip onto in a test that doesn’t ask about intent.

Why the usual mitigations fail

This is worth arguing through rather than just asserting, because the reason both standard defences fail is structural, not incidental. A contractual prohibition on EEA use is a tool for shaping intent — it tells a counterparty what they’re not supposed to do, and creates a remedy if they do it anyway. Article 2(1)(c) doesn’t test intent, so a tool built to manage intent has nothing to act on; the client’s actual use in Spain triggers the provision regardless of what the contract said the client was and wasn’t allowed to do with the deliverable.

Geo-blocking your own service has the same structural mismatch, from a different angle. It controls who can reach your system directly. It does nothing about a customer who received an output outside any geo-blocked environment and later, independently, carries that output somewhere you can’t see and never contracted around. The output has already left your hands by the time the triggering event happens.

The second, quieter expansion

Article 2(1)(g) adds a separate line: the Act applies to affected persons located in the EEA. Under the GDPR, failing every Article 3 test ends the analysis — even if some of the people affected by a business’s processing happen to be in the EEA, that alone doesn’t pull an otherwise out-of-scope business in. Article 2(1)(g)’s wording is genuinely unclear, and White & Case reads it as possibly meaning something different: that an affected person located in the EEA might be able to exercise rights under the Act against a business that passes none of the other territorial tests at all. This is worth flagging precisely as unresolved rather than pushed toward an answer — the source itself doesn’t resolve it, and neither should this piece.

The counter-arguments

None of this is airtight, and the case against reading Article 2(1)(c) this literally deserves its full weight rather than a token mention before moving on. Recitals exist to guide interpretation, and a court taking a purposive approach could plausibly read the “intended to be used” language from Recital 22 back into Article 2(1)(c), resolving the inconsistency in the operator’s favour rather than the literal reading’s. Practically, enforcing the AI Act against a Japanese advertising agency with no EEA presence, no EEA assets, and no EEA representative is a real exercise in fiction — scope on paper and an enforceable judgment are different things, and the gap between them matters. And even where scope does attach, the actual obligation set for a business whose system is genuinely minimal-risk is thin: Article 4 AI literacy, and Article 50 transparency only if the system interacts directly with people. None of the high-risk apparatus this territory usually brings to mind switches on just because Article 2(1)(c)’s trigger fired.

What this means if you sell outside the EU

The honest conclusion sits between the alarming reading and the dismissive one. The exposure is real — the literal text supports it, and the most detailed commentary available concludes there’s no clean way to engineer around it. But for most businesses whose output ends up in the EEA through someone else’s later, unplanned decision, the resulting obligation set is genuinely small. The practical question worth asking isn’t “could we ever technically fall in scope” — for almost any business whose output could travel, the honest answer is probably yes, eventually. It’s which of your systems could ever be high-risk if that trigger fires, since that’s where the actual stakes of this whole analysis live.

The carve-outs that actually help

A handful of exclusions do real work here, separate from the territorial question itself. Article 2(3) excludes AI systems used exclusively for military, defence, or national security purposes. Article 2(6) excludes AI systems and models designed and used solely for scientific research and development. Article 2(8) excludes research, testing, and development of AI systems before they’re placed on the market or put into service — though this exclusion does not extend to testing under real-world conditions, which is treated differently. And Article 2(12) excludes AI systems released under free and open-source licences, except where they are high-risk systems, prohibited systems, or systems subject to the Article 50 transparency obligations. None of these touch the territorial trigger itself, but each one can take a given system out of the analysis before the question of where its output ends up ever needs asking.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base

ISO 42001 Annex A: 38 controls, not 39

Annex A of ISO/IEC 42001:2023 contains 38 controls across nine control objectives, numbered A.2 through A.10. The count matters less than what it’s for: Annex A is a reference set an organisation selects from through its own risk assessment, not a checklist to complete in full.

The answer

38 controls, nine control objectives, running A.2 to A.10:

Objective Covers Controls
A.2 Policies related to AI 3
A.3 Internal organisation 2
A.4 Resources for AI systems 5
A.5 Assessing impacts of AI systems 4
A.6 AI system life cycle 9
A.7 Data for AI systems 5
A.8 Information for interested parties 4
A.9 Use of AI systems 3
A.10 Third-party and customer relationships 3

That totals 38. A.6 is the one objective worth extra care if you’re counting these yourself: it’s the only control objective with a nested, two-level structure — A.6.1’s management-guidance controls, then A.6.2’s system-lifecycle controls beneath it — rather than the flat A.X.Y pattern every other objective uses. It’s the easiest place to miscount by one.

Why the number is the least useful fact about Annex A

Under Clause 6.1.3 of ISO/IEC 42001:2023, organisations select which Annex A controls to implement through their own risk treatment process, document every inclusion and every exclusion with justification in a Statement of Applicability, and may design or implement additional controls beyond Annex A entirely where their risk assessment calls for it. Annex A is a reference set to select from, not an exhaustive checklist to complete.

The practical consequence cuts against the instinct to just implement everything. An organisation that has all 38 controls in place but no documented reasoning behind its selections fails an audit on the Statement of Applicability alone. An organisation with 30 controls implemented, and clear, risk-based justification on record for the eight it excluded, passes. The count of controls implemented was never the thing being tested.

Is Annex B normative?

Annex B gives implementation guidance for each Annex A control — practical detail on how to satisfy a control you’ve selected, not a further set of obligations sitting alongside it. What that means in an audit: an organisation doesn’t have to document or justify its use of Annex B guidance in the Statement of Applicability the way it must for the Annex A controls themselves. Whatever Annex B is formally classified as, it doesn’t carry the audit-tracked weight Annex A does, and that’s the fact worth building a compliance programme around.

On the formal classification itself, the honest position is that the full standard text sits behind ISO’s paywall, and the clearest secondary description available labels Annex B “(normative)” rather than informative. Treat that label as directional rather than fully confirmed, and build your programme around the practical distinction above regardless of how the annex is formally classified.

What the other annexes are for

Annex C sets out potential AI-related organisational objectives and common risk sources — bias, security, misuse, explainability, data quality — functioning as an idea bank when populating the standard’s planning clauses, rather than a set of controls in its own right. Annex D addresses using the AI management system across different domains and sectors, including how the AIMS integrates with ISO/IEC 27001, ISO/IEC 27701, ISO 9001, and sector-specific standards such as ISO 13485 for medical devices.

How this compares to ISO 27001

ISO/IEC 27001:2022‘s Annex A runs to 93 controls across four themes — considerably larger than ISO/IEC 42001:2023’s 38, and organised around a completely different risk object: information security generally, rather than AI specifically.

ISO/IEC 42001:2023 ISO/IEC 27001:2022
Annex A controls 38 93
Number of themes 9 4
Theme breakdown Policy, org, resources, impact assessment, lifecycle, data, information, use, third parties Organisational (37), people (8), physical (14), technological (34)
Risk object AI system behaviour and impact Information confidentiality, integrity, availability

The two annexes genuinely overlap on general governance ground — policy-setting, roles and responsibilities, supplier and third-party management, documented risk treatment discipline. Where ISO/IEC 42001:2023 adds something ISO/IEC 27001:2022 was never built to cover is the AI-specific territory: bias and fairness, transparency to affected people, human oversight of automated decisions, and structured AI impact assessment. An organisation already running an ISO/IEC 27001:2022 ISMS is extending familiar governance machinery into new subject matter, not starting from nothing.

Frequently asked questions

Are all 38 controls mandatory?

No. Annex A is a reference set selected against your own risk assessment, not a checklist every organisation implements in full.

Can we exclude a control?

Yes, provided the exclusion is documented and justified in the Statement of Applicability. An undocumented exclusion is the actual audit risk, not the exclusion itself.

Do we need to implement Annex B?

Not as a separate, SoA-tracked obligation — it’s guidance to help implement the Annex A controls already selected, not an additional layer of controls sitting alongside them.

Is the Statement of Applicability actually audited?

Yes, and it’s the document an auditor checks against far more closely than a raw control count. The SoA is where the risk-based reasoning is supposed to live; an auditor testing an AIMS is testing whether that reasoning holds up, not counting how many of the 38 boxes are ticked.

Does ISO 42001 certification make us compliant with the EU AI Act?

No — they’re separate instruments, and certification under a management-system standard isn’t the same thing as the AI Act’s own presumption-of-conformity mechanism, which attaches specifically to harmonised standards cited in the Official Journal. An ISO/IEC 42001:2023 certificate is strong evidence of AI governance maturity; it isn’t itself a legal finding of AI Act compliance.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base, Workforce, Labour & HR Compliance Reporting

What Is EU GovTech Integration?

Definition

EU GovTech integration is the process of connecting software products, SaaS platforms, and digital workflows with European public-sector systems, registries, identity services, reporting portals, and administrative data exchange frameworks. It can involve APIs, data standards, authentication, digital identity, eDelivery, public registries, compliance reporting, market surveillance systems, procurement platforms, or cross-border public services.

For product, compliance, and SaaS teams, EU GovTech integration is not only a technical connection to a government system. It is a product and compliance capability that requires secure data exchange, interoperability, auditability, legal-role clarity, and operational reliability.

Why EU GovTech Integration Matters

EU GovTech integration matters because public-sector digital infrastructure increasingly shapes how businesses report, verify, authenticate, submit data, and interact with authorities. SaaS products serving regulated customers may need to connect with national portals, EU-level data exchange frameworks, digital identity services, or sector-specific registries.

The EU’s digital government agenda emphasizes interoperability across public administrations. The Interoperable Europe Act entered into force in 2024 and is designed to strengthen cross-border interoperability for digital public services across the EU. Some interoperability assessment and coordination rules started applying from January 2025. (Interoperable Europe Portal)

Core Areas of EU GovTech Integration

Interoperability and data exchange

Teams need to understand the data model, format, protocol, semantics, validation rules, and lifecycle of each public-sector integration. A working API is not enough if the product cannot interpret government responses, handle rejections, or maintain evidence of submission.

Digital identity and authentication

Many GovTech workflows depend on secure identity, authorization, e-signatures, or trusted authentication. The European Digital Identity framework was updated through Regulation (EU) 2024/1183, adopted in 2024, and aims to support EU Digital Identity Wallets for citizens, residents, and businesses. (European Commission)

Cross-border public services

EU GovTech integration can involve cross-border evidence exchange, business registration data, permits, certificates, public procurement, tax, social security, customs, or sector-specific reporting. The Once-Only Technical System is designed to help citizens and businesses provide information to public authorities once and support cross-border data exchange between administrations. (European Commission)

Compliance reporting and audit trails

Many SaaS products integrate with public systems to submit reports, declarations, filings, or compliance evidence. Product teams should design for submission status, error handling, retry logic, timestamps, versioning, user permissions, and proof of delivery.

Third-party and operational dependency management

Government portals and public APIs can change, fail, or behave differently across member states. SaaS teams need monitoring, release tracking, fallback workflows, support playbooks, and clear customer communication when an external public system is unavailable.

Common Implementation Questions

What should teams do first?

Start with a system map. Identify which authority, portal, registry, or framework the product connects to; what data is exchanged; who is legally responsible for the submission; what authentication is required; what evidence must be stored; and what happens if the integration fails.

Is EU GovTech integration only an engineering task?

No. Engineering builds the connection, but product and compliance teams define the workflow, legal assumptions, user permissions, error handling, evidence requirements, support processes, and customer-facing documentation.

What evidence do customers usually request?

Customers may ask for API documentation, data flow diagrams, security controls, authentication methods, audit logs, submission records, uptime expectations, incident procedures, subprocessor details, data residency information, and support escalation paths.

What is the biggest implementation risk?

The biggest risk is treating public-sector integration as a static API task. Government workflows can depend on local rules, changing forms, manual review, unclear error codes, jurisdiction-specific processes, and authority-side outages. The product must support the operational reality, not only the endpoint.

Can a vendor claim EU GovTech compliance?

Use caution. “Fully compliant” is risky unless the vendor defines the jurisdiction, authority, workflow, legal role, data scope, technical standard, evidence model, and date of assessment. Stronger wording explains what the system supports: secure data exchange, identity verification, audit trails, submission monitoring, fallback workflows, interoperability, and customer documentation.

Related Standards and Frameworks

EU GovTech integration often overlaps with the Interoperable Europe Act, European Interoperability Framework, eIDAS and the European Digital Identity framework, Once-Only Technical System, eDelivery, GDPR, cybersecurity controls, public-sector API standards, data exchange specifications, and sector-specific reporting rules.

These frameworks do not create one universal integration pattern. They help teams structure interoperability, trust, data protection, identity, and operational resilience across public-sector workflows.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base

What Is SaaS Data Governance EU?

Definition

SaaS data governance EU is the set of policies, controls, workflows, and technical measures used to manage data in SaaS products operating in or serving the European Union. It covers how data is collected, classified, stored, accessed, shared, protected, retained, deleted, transferred, and documented under EU legal and operational requirements.

For product, compliance, and SaaS teams, SaaS data governance EU is not only a privacy function. It is a product operating model for data quality, accountability, access control, audit readiness, customer trust, and regulatory resilience.

Why SaaS Data Governance EU Matters

SaaS data governance EU matters because EU data rules affect product design, customer procurement, security architecture, vendor management, and go-to-market claims. Teams need to know what data they process, where it is stored, who can access it, which subprocessors are involved, what retention rules apply, and how customers can exercise their rights or meet their own obligations.

The core legal baseline is the GDPR, which has applied since 25 May 2018 and governs personal data processing in the EU. Newer EU data frameworks also matter. The Data Governance Act has applied since September 2023 and focuses on trusted data sharing mechanisms, while the Data Act has applied since 12 September 2025 and creates rules for access to and use of data in connected products and related services, as well as cloud and data processing services.

Core Areas of SaaS Data Governance EU

Data inventory and classification

Teams should maintain a clear inventory of data types, data sources, processing purposes, storage locations, owners, retention periods, and sensitivity levels. This is the foundation for privacy, security, analytics, reporting, and customer due diligence.

Access control and accountability

SaaS products need role-based access control, least-privilege permissions, admin oversight, audit logs, approval workflows, and regular access reviews. Governance should define who can access customer data, under what conditions, and how access is recorded.

Data residency and transfers

EU customers often ask where data is hosted, whether data leaves the EU or EEA, which subprocessors are used, and what safeguards apply to international transfers. Product and infrastructure decisions should support clear answers, not ad hoc explanations.

Retention and deletion

Data governance must define how long different data types are kept, when they are deleted, how deletion is triggered, and whether backups, logs, analytics stores, or third-party tools follow the same lifecycle rules.

Data sharing and interoperability

Modern EU data regulation increasingly focuses on controlled data access, portability, switching, and interoperability. SaaS teams should design export, API, access, and offboarding workflows with governance in mind, not as afterthoughts.

Common Implementation Questions

What should SaaS teams do first?

Start with a data map. Identify all personal, customer, operational, product, analytics, support, billing, and telemetry data. Then map processing purposes, systems, locations, subprocessors, access rights, retention periods, and customer-facing documentation.

Is this only a compliance task?

No. Compliance defines obligations, but product and engineering teams implement the actual controls. Data governance depends on architecture, permissions, logging, UX, APIs, integrations, admin tools, deletion workflows, and customer documentation.

What evidence do customers usually request?

Customers may ask for data processing agreements, subprocessor lists, security certifications, data flow diagrams, retention policies, access control policies, audit logs, transfer safeguards, incident response procedures, backup policies, and deletion processes.

What is the biggest implementation risk?

The biggest risk is fragmented data ownership. If product, engineering, analytics, support, and sales tools all store customer data without a shared governance model, the company may struggle to answer basic questions about access, retention, deletion, and transfers.

Can a vendor claim EU data compliance?

Use caution. “Fully compliant” is risky unless the vendor defines the scope, data types, jurisdictions, legal roles, controls, subprocessors, and date of assessment. Stronger wording explains what the system supports: data inventory, access control, audit trails, retention workflows, deletion workflows, transfer documentation, and customer governance evidence.

Related Standards and Frameworks

SaaS data governance EU often overlaps with GDPR, Data Governance Act, Data Act, ISO/IEC 27001, ISO/IEC 27701, SOC 2, NIST Cybersecurity Framework, cloud security controls, privacy management systems, and internal data governance policies.

These frameworks do not replace legal analysis, but they help structure controls, evidence, ownership, monitoring, and audit readiness.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base

Your First 90 Days of DORA — An IT Controls Implementation Roadmap for Fintech Startups

DORA is now live across the EU, setting uniform rules for ICT risk management, incident reporting, testing, and third‑party risk for financial entities — and it applies from 17 January 2025. If you’re a Fintech startup operating in Europe, your next 90 days determine whether you can withstand audits, secure partnerships, and continue scaling in regulated markets. Here is a pragmatic, regulator‑aligned plan you can execute immediately.

Who this roadmap is for

  • CTOs, Heads of Product/Engineering, CISOs in EU Fintechs that fall under DORA’s scope (e.g., payments, e‑money, investment, insurance intermediaries, crypto‑asset service providers where applicable).
  • Founders preparing for bank partnerships, licensing, or entering multiple EU markets.

DORA in 60 seconds — what you must operationalize

  • ICT risk management — governance, policies, controls, monitoring.
  • Major ICT incident reporting — classification thresholds, initial/intermediate/final reporting timelines, harmonised templates.
  • Digital operational resilience testing — from basic testing to TLPT, aligned with TIBER‑EU and the TLPT delegated act.
  • ICT third‑party risk — Register of Information (RoI) templates, contract clauses, subcontracting, and CTPP oversight.
  • Information sharing and sector coordination — lex specialis vis‑à‑vis NIS2 for financial entities.

The 90‑day sprint at a glance

  • Days 1 — 30: Foundation and gap assessment
  • Days 31 — 60: Control implementation and monitoring
  • Days 61 — 90: Proving resilience — drills, testing posture, audit evidence

Day 0 — Prerequisites and success criteria

  • Name a senior accountable owner (management body responsibility).
  • Define success metrics: mean time to detect/respond (MTTD/MTTR), backup success rate, patch SLAs, vendor coverage, incident drill timings.
  • Set up an “evidence library” from day one: store policies, approvals, scans, monitoring screenshots, reports, tickets.

Days 1 — 30: Build your foundation

1) Governance and risk

  • Approve an ICT Risk Management Policy and control standard; adopt a risk taxonomy aligned to DORA.
  • Map business services to “critical or important functions” (CIFs) and define RTO/RPO targets.
  • Perform a gap assessment against ISO 27001, NIST CSF, and SOC 2 to prioritize near‑term controls.

2) Incident reporting playbook — align to RTS/ITS timelines

  • Create a one‑page escalation matrix from “detection” to “classification” to regulator notification.
  • Implement timers and checklist for the DORA time limits: initial notification within 4 hours of classification and within 24 hours of detection; intermediate at 72 hours; final within 1 month. Rehearse once in the first 30 days.
  • Prepare data fields required by the standard forms and templates (ITS) to avoid scrambling during an incident.

3) Register of Information (RoI) — start early

  • Stand up your RoI using the official templates (providers, services, CIF linkage, locations, data types, subcontractors, exit plans).
  • Identify providers using LEI or EUID as permitted; capture both if available.
  • Set a weekly governance slot to enrich and validate the RoI — it becomes supervisory data and feeds CTPP oversight.

4) Third‑party contracting and procurement controls

  • Insert DORA‑aligned clauses: audit/access rights, incident notification, subcontracting conditions, exit/termination, data location/sovereignty, resilience KPIs.
  • Flag critical/important functions and define pre‑approval for any subcontracting chain.

5) Proportionality and simplified framework

  • If you qualify for the simplified ICT risk management framework, document justification and scope, but do not skip core controls (asset inventory, access control, backup/restore, logging, vulnerability management).

Deliverables by Day 30

  • Board‑approved ICT policy and risk appetite
  • Gap assessment and 90‑day remediation plan
  • Incident reporting runbook and on‑call rota
  • RoI v0.9 completed for top 10 vendors
  • Contract addendum template for DORA clauses

Days 31 — 60: Implement controls and monitoring

1) Core controls that stick

  • Asset inventory and CMDB; access governance (RBAC, MFA, joiner‑mover‑leaver), secure default configurations.
  • Logging and monitoring: centralize application and infrastructure logs; set alerts for availability, integrity, and authentication events.
  • Backup and recovery: test restores for systems supporting CIFs; evidence RTO/RPO drills.
  • Vulnerability management: 14–30 day SLA by severity; change management tied to risk.
  • SDLC controls: code scanning, dependency risk, secrets management, pre‑prod testing.

2) Incident readiness — drill against the clock

  • Run a tabletop exercise with the 4‑hour/24‑hour/72‑hour/1‑month reporting cadence; fill the templates you’ll actually submit.
  • Add communications paths to your competent authority’s portal/process.

3) RoI hardening and supervisory readiness

  • Extend RoI coverage to 100% of ICT third‑party arrangements; link each to CIFs and exit strategies.
  • Validate provider identifiers (LEI/EUID) and subcontractor chains for critical services.

4) Concentration and CTPP awareness

  • Identify cloud/SaaS concentration, region dependencies, and portability gaps.
  • Understand the 2025 CTPP designation process so you can respond to oversight implications downstream.

Deliverables by Day 60

  • Monitoring dashboards and alert runbooks
  • Evidenced restore test and failover notes
  • Vulnerability backlog burn‑down; patch metrics
  • RoI v1.0 complete with data quality checks

Days 61 — 90: Prove resilience and close gaps

1) Testing posture — from basic to TLPT

  • Document your testing scope: vulnerability scans, config reviews, incident drills, supplier failover tests.
  • Decide if and when you may fall under Threat‑Led Penetration Testing (TLPT) and align future steps with the TLPT delegated act; track the Eurosystem TIBER‑EU alignment for deliverables and approach.

2) Third‑party resilience and exit

  • Walk through an exit drill for one critical SaaS: data export, re‑platform steps, and roll‑back.
  • Verify contractual rights for audits, on‑site visits, and incident cooperation.

3) Audit evidence and metrics

  • Consolidate a 12‑item evidence pack (see checklist below) and lock your KPI baselines (MTTD/MTTR, backup success, patch SLAs, incident drill timings).
  • Schedule a mock supervisory Q&A using your RoI and incident templates.

Deliverables by Day 90

  • Testing strategy and TLPT readiness memo
  • Executed third‑party exit rehearsal notes
  • Complete evidence library and KPI baseline

Control checklist — what supervisors will expect to see

  • Governance: management body minutes, risk appetite, policy approvals.
  • Risk: CIF mapping, risk register, treatment plans.
  • Assets & access: inventories, access reviews, MFA coverage, admin hardening.
  • Monitoring: logging scope, alert rules, SIEM/SOAR playbooks.
  • Data protection: encryption, key management, secure backups, restore proofs.
  • Change & vuln: change tickets linked to risk, scan results, patch reports.
  • Incident: classification matrix, on‑call rota, reporting templates, drill evidence.
  • Third‑party: RoI, due diligence packs, DORA clause addenda, subcontracting disclosures.
  • Testing: test plans, results, fixes, and TLPT roadmap where applicable.
  • Metrics: MTTD/MTTR, RTO/RPO tests, vendor coverage, concentration analysis.

Mapping DORA pillars to popular frameworks

DORA requirement ISO 27001 (2022) NIST CSF 2.0 SOC 2 (TSC)
ICT risk management A.5, A.8, A.5.36, A.5.23, A.5.30 Identify, Protect Security, Availability
Incident mgmt & reporting A.5.24, A.5.25, A.5.29 Detect, Respond Security, Availability
Digital resilience testing A.8.8, A.8.9, A.8.16 Detect, Respond Security
ICT third‑party risk & RoI A.5.19, A.5.20, A.5.21 Identify, Protect Security, Availability
Info sharing & coordination A.5.7 (relevant) Identify, Respond Security

Use this mapping to translate existing audits into DORA evidence and to spot gaps quickly.

Incident reporting — the operational details that matter

  • Classify and timebox: initial notification within 4 hours of classification and within 24 hours after detection; intermediate in 72 hours; final within 1 month. Build timers into your on‑call flow.
  • Populate the harmonised forms: root cause hypothesis, impact, mitigations, cross‑entity effects, and external dependencies per the ITS.
  • Align overlaps with NIS2 where applicable, noting DORA’s lex specialis role for financial entities.

ICT third‑party risk — RoI, contracts, and CTPP oversight

  • Maintain the RoI at entity/sub‑consolidated/consolidated levels, using the official templates; ensure CIF linkage, data locations, and subcontractor chains are captured.
  • Use LEI or EUID for provider identification as per the Commission’s position clarified by supervisors; validate identifiers early to avoid rework.
  • Expect authorities to collect RoIs and feed ESAs’ 2025 criticality assessments — submit clean, deduplicated data on request.
  • Keep contract conditions aligned with the delegated RTS on third‑party policy and subcontracting expectations.

TLPT and testing — right‑sizing for startups

  • Not all entities will immediately be in scope for TLPT, but you should maintain a forward plan aligned with the TLPT delegated regulation and TIBER‑EU updates.
  • Prioritize basic testing with real fixes: restore tests, failover, key compromise drill, supplier outage simulation, and a red/blue tabletop.

Evidence library — what to file for audits and partnerships

  • Board approval of ICT policy and risk appetite
  • CIF map, RTO/RPO, data flow diagrams
  • Incident classification matrix, drill decks, and filled templates
  • Monitoring dashboards, alert rules, and sample tickets
  • Backup and restore logs, failover test reports
  • Vulnerability scans, patch cadence reports
  • RoI export (current), vendor risk files, contract addenda
  • TLPT readiness memo and annual test plan

Interplay with NIS2 — clarity for multi‑regulated firms

For financial entities within scope of DORA, DORA operates as a sector‑specific lex specialis for cybersecurity measures and incident reporting; do not duplicate obligations under NIS2 where DORA applies. Ensure your internal compliance matrix reflects this to avoid double reporting.

Common pitfalls to avoid

  • Treating RoI as a procurement spreadsheet — it is a supervisory artifact with strict taxonomy and identifiers.
  • Reusing US‑centric SOC 2 controls without EU regulatory mapping — you’ll miss incident reporting and subcontracting depth.
  • Underspecifying exit and portability for critical SaaS — regulators will ask for evidence.
  • Practicing incident response without the real RTS/ITS time checks and forms.

What to do next — a 2‑week quick start

1) Approve ICT policy and risk appetite; name accountable owner. 2) Stand up RoI with top vendors and correct identifiers; schedule weekly data quality checks. 3) Run a 60‑minute tabletop using the 4h/24h/72h/1‑month cadence and fill the templates once end‑to‑end. 4) Lock backup/restore tests for systems supporting CIFs and record evidence. 5) Issue DORA addendum to critical supplier contracts.

Summary

  • DORA is in force EU‑wide — the next 90 days are about operationalizing controls, not drafting policies.
  • Nail three things early: a working incident playbook with RTS/ITS timings, a regulator‑grade RoI with correct identifiers, and a tested recovery capability.
  • Right‑size testing now and plan for TLPT alignment; keep contracts and subcontracting under tight control, and maintain clean evidence for supervisors and partners.
Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base, What happened with...

Data Architecture — Designing a Data Model for the Digital Product Passport: A Step‑by‑Step Walkthrough

The EU’s Digital Product Passport (DPP) is moving from policy into implementation — and the quality of your data model will make or break compliance, interoperability, and product velocity. For European tech leaders and policy stakeholders, this guide translates legal intent into a concrete, scalable data architecture that satisfies regulators while creating real business value.

Below, you’ll find a pragmatic, step‑by‑step approach to model design — from identifiers and lifecycle events to access control, verifiability, and systems integration across GovTech and LegalTech ecosystems.

TL;DR — What you will get right if you follow this guide

  • A standards‑aligned data model that maps to ESPR requirements and industry vocabularies.
  • Clear separation of public vs restricted data — designed for privacy, IP protection, and market surveillance.
  • Interoperability by default — using GS1 Digital Link, EPCIS 2.0, JSON‑LD, and W3C Verifiable Credentials.
  • A practical rollout plan — start with a Minimal Viable Passport and iterate per product category.
  • Governance you can operate — roles, stewardship, and controls that auditors and regulators can trust.

What is the Digital Product Passport — in plain terms

A Digital Product Passport is structured product information that travels with the product across its lifecycle — design, manufacturing, distribution, use, repair, and end‑of‑life — to support circularity, transparency, and compliance across the EU. It is accessed via a data carrier such as a QR code and exposes both public information for consumers and controlled information for professionals and authorities.

In simple terms: it’s a durable, scannable “file” about your product — what it is, what it contains, how it was made, how to repair it, how to recycle it, and whether it meets EU rules.

Design principles for a DPP‑ready data model

  • Interoperability first — align to open standards and shared vocabularies to prevent vendor lock‑in.
  • Verifiable and tamper‑evident — support digital signatures and credential formats to prove authenticity.
  • Least‑privilege transparency — publish what’s needed publicly while safeguarding sensitive or personal data.
  • Lifecycle completeness — model events, not just master data, to cover circularity and compliance needs.
  • Operationally realistic — keep payloads lightweight for scanning performance and offline access.
  • Governance by design — embed stewardship, quality rules, lineage, and audit into the model.

Step‑by‑step walkthrough

Step 1 — Define scope, stakeholders, and data audiences

Identify required product categories and implementation timelines, then map who needs to see what.

  • Consumers — product claims, safety instructions, care/repair, environmental info.
  • Professionals (repairers, recyclers) — parts, schematics, disassembly, hazard notes.
  • Authorities — conformity data, certificates, traceability events, recalls.
  • Partners (suppliers, logistics) — materials, batches, lifecycle events.
  • Internal teams — master data, BOM, quality, EHS.

Create a RACI per attribute group to make ownership explicit.

Step 2 — Choose identifiers and URIs

Consistency starts with identity:

  • Product master — GTIN, brand owner, model/variant codes.
  • Serialized item — SGTIN/serial or equivalent per category; batch/lot where applicable.
  • Facilities — GLN for sites; ISO country codes; internal plant IDs.
  • Digital Link URI — the QR‑encoded entry point that routes to the right passport view.
  • Optional decentralized identity — DIDs for issuers/holders if you adopt verifiable credentials.

Rule of thumb: one canonical identifier per entity, with stable cross‑references and versioning.

Step 3 — Define canonical entities and attributes

Create a small, durable set of entities that cover 80–90% of scenarios:

  • Product, ProductVariant, Item (serialized), Batch
  • Material, Component, Substance
  • EconomicOperator (manufacturer, importer, distributor, repairer, recycler)
  • Facility (site), Equipment
  • Certification, Declaration, Assessment
  • Document (attachments), Image, Instruction
  • LifecycleEvent (EPCIS), Shipment, Return, Repair
  • EnvironmentalFootprint, Energy, Durability, Repairability
  • AccessPolicy, Credential, Consent
  • DataSource, DataContract, DataLineage

Keep attribute names human‑readable and map them to standard vocabularies via JSON‑LD contexts.

Step 4 — Model lifecycle events using EPCIS 2.0

DPP is not just static facts — it’s what happened and where. Use event modeling to represent:

  • Commissioning, transformation, aggregation, shipping/receiving, disaggregation
  • Business step, disposition, readPoint (where), bizLocation (who), time, source/destination
  • Link events to items, batches, and certificates to form provenance chains

This powers recall traceability, warranty, and circularity analytics.

Step 5 — Capture compliance and sustainability data

Create structured sub‑documents for:

  • Conformity and certificates — CE marking basis, harmonized standards, validity periods
  • Restricted substances and material declarations — with thresholds and references
  • Durability and repairability indices — with test protocols and scoring methods
  • Energy performance and environmental footprint — method, system boundary, version, verifier
  • Safety and instructions — multilingual, accessible formats, versioned

Model both the value and the methodology — auditors will check how you obtained the number.

Step 6 — Design access tiers and policy rules

Separate the passport into tiers:

  • Public — safe for consumers and marketing claims
  • Restricted — professional access (after role or credential check)
  • Confidential — authority access or NDAs, never public

Use role‑ and attribute‑based policies: role = “authorized repairer” AND device.category = “electronics” AND brand = “X”. Link policies to attributes at the field level, not just the document.

Step 7 — Data quality, lineage, and governance

Treat DPP as an MDM program with external exposures:

  • Golden sources and data contracts per attribute domain
  • Quality rules — required fields per category, allowed value sets, freshness SLAs
  • Lineage — where data came from, when, and under which contract
  • Stewardship — named owners, escalation, and change control

Step 8 — Storage and persistence patterns

Use polyglot storage to balance scale and flexibility:

  • Relational for master data and governance
  • Document store for passport payloads and localized content
  • Graph for relationships (items ↔ components ↔ suppliers)
  • Object storage for documents and media
  • Event store for EPCIS events, partitioned by time and GTIN

Keep public payloads small — retrieve restricted details on demand behind policy checks.

Step 9 — APIs and exchange formats

Prioritize JSON‑LD for semantics, and design stable, versioned APIs:

  • Read operations — per GTIN, per serial, per Digital Link URI
  • Event ingestion — EPCIS JSON events with schema validation
  • Credential endpoints — issue, verify, revoke W3C Verifiable Credentials
  • Webhooks — recall notices, certificate expiries, non‑conformance alerts

Use idempotency keys and ETags to support caching and offline behavior.

Step 10 — Security, privacy, and legal alignment

Operationalize EU Data Compliance & Strategy:

  • Data minimization — publish only what each audience requires
  • GDPR — personal data avoidance; if unavoidable, define legal basis, retention, and DSAR handling
  • Trade secrets — field‑level policies and non‑extractable rendering where possible
  • Integrity — digital signatures, tamper‑evident logs, clock synchronization
  • Auditability — immutable event logs and verifiable credential proofs

Step 11 — Data carrier and user experience

QR scanning must be fast, resilient, and predictable:

  • Digital Link routing with locale negotiation and device‑appropriate views
  • Offline snapshots for critical safety information
  • Graceful degradation and long‑lived redirects if product URLs change
  • Accessibility — WCAG compliance, simple labels, clear disclosures

Step 12 — Migration and incremental rollout

Start with a Minimal Viable Passport for a single category, then expand:

  • Phase 1 — product identity, public attributes, basic certificates
  • Phase 2 — BOM summary, repairability, recyclability guidance
  • Phase 3 — lifecycle events, professional data, authority interfaces
  • Phase 4 — verifiable credentials, partner integrations, analytics

Reference schema — practical examples

Product master (JSON‑LD)

{
“@context”: [
“https://schema.org/”,
{ “gs1”: “https://gs1.org/voc/” }
],
“@type”: “Product”,
“gtin”: “04012345123458”,
“name”: “Model X — Smart Washer 700”,
“brand”: “Acme”,
“model”: “SW-700”,
“category”: “household-appliance”,
“manufacturer”: {
“@type”: “Organization”,
“name”: “Acme Appliances GmbH”,
“leiCode”: “529900T8BM49AURSDO55”
},
“digitalLink”: “https://id.acme.example/p/04012345123458”,
“dpp”: {
“version”: “1.0.0”,
“status”: “active”,
“publicAttributes”: [
“safetyInstructions”,
“repairabilityScore”,
“energyLabel”
],
“restrictedAttributes”: [
“disassemblySteps”,
“partsCatalog”
]
}
}

Serialized item and lifecycle link

{
“itemId”: “04012345123458.1234567890”,
“scheme”: “sgtin”,
“batch”: “B24-10-001”,
“manufactureDate”: “2025-09-22”,
“facilityGLN”: “1234567890123”,
“events”: [
“urn:epcis:event:3f4a…”,
“urn:epcis:event:7c21…”
]
}

EPCIS 2.0 event (transformation example)

{
“type”: “TransformationEvent”,
“eventTime”: “2025-09-23T10:11:12Z”,
“inputEPCList”: [“urn:epc:id:sgtin:4012345.012345.1234567890”],
“outputEPCList”: [“urn:epc:id:sgtin:4012345.012345.2233445566”],
“bizStep”: “urn:epcglobal:cbv:bizstep:assembly”,
“disposition”: “urn:epcglobal:cbv:disp:in_progress”,
“readPoint”: { “id”: “urn:epc:id:sgln:1234567.00001.0” },
“bizLocation”: { “id”: “urn:epc:id:sgln:1234567.00001.0” },
“ilmd”: {
“acme:processVersion”: “2.7.1”,
“acme:operatorId”: “EMP-9982”
}
}

Certificate as a Verifiable Credential

{
“@context”: [
“https://www.w3.org/2018/credentials/v1”,
“https://example.eu/credentials/conformity/v1”
],
“type”: [“VerifiableCredential”, “ConformityCertificate”],
“issuer”: “did:example:nb-12345”,
“issuanceDate”: “2025-10-01T00:00:00Z”,
“credentialSubject”: {
“gtin”: “04012345123458”,
“category”: “household-appliance”,
“standards”: [
{ “ref”: “EN 50564”, “version”: “2011” }
],
“validUntil”: “2028-10-01”
},
“proof”: {
“type”: “Ed25519Signature2020”,
“created”: “2025-10-01T00:00:00Z”,
“verificationMethod”: “did:example:nb-12345#keys-1”,
“proofPurpose”: “assertionMethod”,
“jws”: “eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9…”
}
}

Access policy — field‑level control (conceptual)

{
“resource”: “dpp:04012345123458”,
“rules”: [
{
“effect”: “allow”,
“fields”: [“public.*”],
“condition”: “audience == ‘public'”
},
{
“effect”: “allow”,
“fields”: [“restricted.disassemblySteps”, “restricted.partsCatalog”],
“condition”: “role in [‘authorized_repairer’, ‘manufacturer’]”
},
{
“effect”: “deny”,
“fields”: [“confidential.*”],
“condition”: “audience != ‘authority'”
}
]
}

Public vs restricted vs confidential — quick mapping

Tier Examples Who sees it Risk if public
Public safety, energy label, repairability score, materials summary everyone low
Restricted parts exploded views, exact disassembly steps, batch events repairers, recyclers, partners moderate (IP, safety misuse)
Confidential supplier pricing, QA non‑conformities, internal SOPs authorities, auditors, selected internal high (trade secrets)

 

Centralized vs federated vs decentralized — which pattern fits

  • Centralized — one platform stores and serves all passport data. Simple to operate, faster rollout, but higher concentration risk and potential vendor lock‑in.
  • Federated — each operator hosts its data; a resolver composes a passport view. Better data sovereignty and scalability; requires strong discovery, policy, and caching.
  • Decentralized credentials — authoritative facts are issued as verifiable credentials and verified at read time. Strong authenticity and portability; more complex key and revocation management.

Most enterprises start centralized for speed, then evolve to federated with credentials for high‑assurance facts.

KPIs and controls that matter

  • Coverage — % of SKUs with DPP, % of serialized items linked
  • Freshness — median passport update latency after product change
  • Data quality — required field completeness, schema conformance rate
  • Access — P95 QR resolve time, authorization error rate, offline fallback rate
  • Assurance — % of credentials verified, audit findings closed on time
  • Circularity — repair rate uplift, components reused, EoL capture rate

Common pitfalls — and how to avoid them

  • One giant document — split into composable resources with field‑level policies.
  • Ignoring standards — map early to GS1 Digital Link and EPCIS to avoid rework.
  • Over‑publishing — leakage of IP or personal data; apply tiering and minimization.
  • No lineage — auditors will ask “who said this, when, and under which method?”
  • Heavy payloads — slow scans and timeouts; keep public views lightweight with just‑in‑time expansion.

Implementation roadmap — 120‑day starter plan

1) Weeks 1–3 — scoping and discovery: categories, stakeholders, attributes, risks. 2) Weeks 4–6 — data model v1, JSON‑LD context, identifier strategy, policies. 3) Weeks 7–9 — API and resolver services; basic Digital Link and QR routing. 4) Weeks 10–12 — MDM onboarding, content localization, certificate ingestion. 5) Weeks 13–16 — pilot launch on one product line; monitoring and KPI baseline. 6) Weeks 17–18 — hardening: access controls, signatures, audit logging. 7) Weeks 19–20 — rollout to second category; add lifecycle events and pro data.

FAQ — quick answers for decision‑makers

  • Do we need to serialize every item? Not always — some categories work at batch/lot level, but serialization improves traceability and repairs.
  • Is a blockchain required? No — verifiable credentials and signed events provide authenticity without on‑chain storage.
  • Can we reuse product content systems? Yes — integrate via data contracts; treat DPP as a governed product data exposure.
  • How do we handle recalls? Model authority‑visible lifecycle events and push recall notices via webhooks and public flags.
  • What about GDPR? Design to avoid personal data in DPP; where unavoidable, apply legal basis, retention, and DSAR processes.

Summary — make compliance your product advantage

A robust DPP data model is a strategic asset — not just a regulatory chore. Center it on interoperable identifiers, lifecycle events, verifiable claims, and audience‑appropriate access. Start small with a Minimal Viable Passport, standardize your entities and APIs, and build governance that stands up to regulatory and market scrutiny.

Get these foundations right and you’ll ship faster, satisfy EU expectations, and unlock real circularity value — all while aligning GovTech and LegalTech integration with your broader Data Compliance & Strategy.

Posted by admin in Digital Operational Compliance & EU AI Act Knowledge Base, What happened with...