Skip to content

Wargo French Singer to join with DWT, effective January 1, 2027. Learn more.

DWT logo
People Services Insights
About Offices Careers
Search
People
Services
Insights
About
Offices
Careers
Search
Webinars
Technology + Privacy & Security

"Preparing for the EU Cyber Resilience Act: A Roadmap for U.S. Companies," Davis Wright Tremaine Webinar

11.12.26
Share
Print this page

The European Union's Cyber Resilience Act (CRA) is reshaping cybersecurity compliance for companies that develop and sell connected products and software. For U.S. companies, the CRA's reach can extend well beyond Europe: Products made available on the EU market may be subject to new requirements governing secure product design, cybersecurity risk assessments, vulnerability management, incident reporting, and security updates, among others. The compliance clock is already running, with the CRA's vulnerability and incident reporting obligations taking effect in September 2026, ahead of the broader requirements that apply beginning in December 2027.

In this webinar, Andrew Lewis from Davis Wright Tremaine and Tobias Rothkegel from Osborne Clarke will explain how the CRA applies to U.S. companies and what organizations should be doing now to prepare. The discussion will also examine the CRA's rapid incident and actively exploited vulnerability reporting requirements, including its 24- and 72-hour reporting framework, as well as how the CRA intersects with other European regulations. The program will provide U.S. companies with a practical roadmap for moving from CRA scoping to operational readiness before the broader regime takes effect.

Speakers

  • Andrew Lewis, Counsel, Davis Wright Tremaine
  • Tobias Rothkegel, Partner, Osborne Clarke

Click to download presentation slides.

(On-demand versions are not eligible for CLE credit.)


Webinar Transcript

The following transcript was created using AI-assisted speech recognition and formatted for publication. Readers should be aware that automated transcription may not capture every statement with complete accuracy.

Andrew Lewis: Thank you all for joining us today. Today we're going to talk about the EU Cyber Resilience Act, discuss sort of what it means for U.S. companies. Slides will be available. Today's focus is not simply on what the CRA says, but we want to explain where companies are likely to encounter both difficult classifications and implementation questions.

So, if you have questions on anything, you can drop them in the chat, and we will reserve some time at the end to answer them. This webinar is eligible for CLE credit, so please be sure to fill out the attendance verification form that has been provided via email. It will also be re-shared in the chat for our attendees. You'll need to listen aloud for the CLE code, which we'll read later in the presentation. So with that, let's go ahead and get started. So, I am Andrew Lewis, Cybersecurity Counsel here at Davis-Wright Tremaine. I am joined by Tobias Rothkegel. Tobias, do you want to introduce yourself?

Tobias Rothkegel: Sure. Thanks, Andrew. Thanks for the invite. Very glad to be here. Good morning, everyone. My name is Tobias, I'm a German-qualified lawyer and partner at Osborne Clark. Here in the Hamburg office, I had our cybersecurity practice here in Germany, and for those who don't know Osmo Clark, we are a UK-based, full-service law firm, but we have very much, focus on… on tech, data, cyber, so we work a lot with, Davis Wright, Tremaine, and we have a lot of, also clients on the On the West Coast, we're very accustomed to the… to the issues US companies face when entering… entering Europe.

Andrew Lewis: Thanks, Tobias. And yes, we are approaching this sort of jointly from a US and EU perspective, because I think it's helpful to have sort of both views on what this regulation means. So thank you, Tobias, for joining, and thank you for joining in, I assume, is the middle of the night your time, so really appreciate that.

Tobias Rothkegel: It's not as… it's 7pm, it's not as bad.

Andrew Lewis: Still, still appreciated it. Okay. So we're gonna do just, we're gonna do a pretty comprehensive, so, we'll dive into what CRA requires, but we want to give an overview for those who are unfamiliar with it, or are only familiar with it from the incident response portion that went active this year. So CRA is broader than that, and so, you know, bear with us while we sort of talk about the overall what is the CRA? So, the Cyber Resilience Act is a binding EU law which establishes broad, minimum cybersecurity requirements related to both the design, development, and production of products with digital elements made available on the EU market.

That's sort of the definition and what triggers CRA application. There's a lot to unpack there, because every one of those words has a definition associated with it. And we'll dive into that a little bit here before we dive into what it actually requires. One thing to think about, though, is that the goal of the law is really to establish minimum cybersecurity standards across Europe. So if you think about that while we sort of talk about everything else for the rest of the day, it'll make a lot of sense, that the idea is to put secure products into the European market. First, one of the sort of nuances here is that this is actually an EU law.

It is a regulation, not a directive, and so it doesn't require individual member state adoption. that's about as much as I can describe on that, so Tobias, can you sort of describe the difference between the laws, the directives? We're dealing with a lot of them here, and I don't think most, at least I don't, quite understand the distinction.

Tobias Rothkegel: Yeah, sure. So, I think you already hit the nail on the head. So, the major difference between an EU regulation and an EU directive is that the directive in itself has no binding effect for companies or civilians. It only has binding effect for the member states, because they need to transpose the requirements of the directive into a national law, which then means, every member state has its own national law, which can deviate to some point.

So, for instance, when we're talking about, cybersecurity regulation, NIST2, which I guess most of you have also heard of, which is another piece of very important cybersecurity legislation in Europe, is a directive, because, the… So the rationale behind this was that the member states shall have certain autonomy to, tailor the provisions to the realities of the market in those. member states.

And the CRA, on the other hand, is a regulation which is directly applicable, doesn't need to be transposed to national law, so it has a direct binding effect for Companies and civilians, with the rationale being that it's a more product-focused, regulation, which doesn't need to, kind of cater towards, national, differences, so to speak. So, the CRA is applicable live directly how it is.

There will be accompanying national laws, because the, national law, the member states need to appoint, supervisory authorities, for the enforcement, they need to. do some administrative, framework around this, so there will be some sort of national accompanying law, like, for the GDPR, as well, so there are also national laws kind of flanking the GDPR also in the member states, but it will never be as relevant for, for instance, US companies, as the CRA in itself is. So the CRA is kind of like the… the major attraction here.

Andrew Lewis: Perfect. Really helpful. Thank you for that, that succinct explanation of that. So the other aspect here, and Tobias, I'm going to want to get your thoughts on this as well, the product with digital elements, that's sort of the trigger on what this covers, what the CRA covers, and that's software or hardware and their remote data processing solutions, And it broadly applies to most hardware and software that can be connected to another device or network. And, Tobias, what I want to ask you about is, I think everyone's clear that products that are sold off the shelf are covered. Hardware that is sold in the EU is covered. websites or not.

There's this gray area in between on SaaS, and I want to get your thoughts on what sort of… would… would or would not bring SAS into the regulation.

Tobias Rothkegel: Sure. So, I think the… for me, at least, the easiest way to differentiate when SaaS is actually CRA relevant and when it is not is the comparison with the website, as you said. Like, website is not, is not covered by the CRA, because you just open your browser and visit it. And if your SaaS service works basically in the same… in the same manner, where you just open your browser and, access the the service, then it is not covered by the CRA, because it is intended to be covered by NIST2. Right? If you, then fulfill the company size thresholds. So, like, SARS in its pure form is intended to be regulated by NISTU and not by the CRA.

But then you have, like, those hybrid solutions. Like, for instance, if you have, like, a client software you need to install on your machine at home to access a SaaS service, then it becomes… CRA relevant. Then it is a product with digital elements, because you have Some sort of piece of code, or maybe it can also be a piece of hardware that is, introduced into your sphere, so… if you want to frame it like that. And then you kind of invite certain risks into your sphere, into your IT infrastructure, and then it becomes CRA relevant.

So, for me, like, the rule of thumb is like, pure online-based SAS, not covered, SAS where I need a client software installed on my machine, or can be also an app on my phone, then it is taught by the CRA.

Andrew Lewis: Again, very helpful, and I think as, as… listeners are noting this is a complicated question that's going to require a lot of analysis, so I think that even that threshold question is going to require a lot of work from legal and product teams to determine whether or not the… the device or product is even covered by the CRA. So the other trigger here is that it is made available on the EU market. And so the CRA applies when covered products are supplied for distribution or use on the EU market in the course of a commercial activity. And that can be for free or for charge, it just has to be sort of related to a commercial activity.

So free doesn't necessarily mean outside of commercial activity, because a free download, for instance, could still be made available as part of some other paid platform, monetization service. So it being… it in and of itself being free doesn't necessarily mean that it isn't applicable. The other one here that I think is kind of… Lumped together is open source. That is also not a blanket exception.

So free and open source software, supplied outside of a commercial activity is not… subject to the CRA, but if you sort of trigger that commercial activity as part of it, either you're providing it as part of a commercial activity, or you as a manufacturer are embedding open source software into a product that is then sold as part of commercial activity, both of those would fall under the CRA. So, another just important caveat that just because it's free, just because it's open source, doesn't mean that it gets a blanket exception. It would require some legal nuance and discussion on whether or not, the CRA applies.

And just last point here, timeframe. Most obligations trigger… are triggered in December of 2027, but incident reporting began, back in September of this year. So we'll talk about the incident reporting piece. That was sort of the only piece that was activated early. Okay, we talked a little bit about this earlier, but we just want to sort of discuss what's different about the CRA, because it is, both it in its application as a law and the way that it is framed behaves differently than a lot of other European regulations and directives. So again, back to the point that it is a broad, uniform level of cybersecurity across the European Union.

That's its goal. It is also product-focused. Unlike other… laws like NIST2, which are sort of sector-specific, and have thresholds in order to meet. This is really focused on the individual product, so it applies to covered hardware and software products, and the integrated remote data processing solutions, as Tobias mentioned, but it's on a product-by-product basis. not a company-by-company or even sector-by-sector basis. So you're gonna have to analyze does CRA apply to this individual product pretty much every time you launch a product or change a product, which we'll get to in a second.

Again, the EU market trigger, so this is somewhat different than other things, but basically is only applicable if you are triggering release of the product in the EU market during the course of the commercial activity. That's related to other European laws, so not completely new, but still a different trigger point than we're normally seeing. There are also these role-based obligations, so you see different obligations applied to different parts of the, sort of, supply chain.

So, the primary obligations apply to manufacturers, and there are also responsibilities to the importers, distributors, and authorized representatives, which I think generally speaking, U.S. companies are mostly going to be manufacturers, but those other entities, I think, are going to play a role in how U.S. companies institutionalize the CRA, because they may have requirements pushed down onto the U.S. companies, because, like, through contract and whatnot. So you'll see sort of an interaction between U.S. companies who are making products and their European, counterparts who are selling them.

The other interesting part about this that I… we don't often see is that there are really lifecycle obligations built into the CRA, so it… regulates from, sort of cradle to grave, right? Design and development, all the way through the vulnerability handling, reporting, updates, and post-market action that requires the manufacturer to provide some level of certainty on security throughout the life of the product. And to consider that while designing the product.

So, again, something new that we're seeing here, that should be taken into consideration. because… Companies really do need to think about What does this product look like in a couple of years, and what kind of security are we going to be able to provide for it in the long term? You know, one practical difficulty that that means is that no one department is going to own the obligations here. Legal is obviously going to be involved, product design, product security are going to be involved. A variety of engineering teams are gonna be involved, and… More importantly to this.

Lifecycle obligation is that you're going to need to reassess a lot of this stuff. So there needs to be… sort of triggers built in to re-engage those groups throughout the life cycle of the product, which I think is… it exists, but hasn't been as formalized of a requirement. Okay. I'm gonna just click through all this, because this is… There we go. Tobias, do you want to, discuss how U.S. companies can be subject to the CRA?

Tobias Rothkegel: Sure, sure, thanks, thanks, Andrew. And I think a very important aspect to take away from this today is that, in essence, the CRA is a market entry requirement for the EU. So, it is kind of disguised as a product regulation, or as a product security regulation, but in essence, it basically says if you don't comply, you cannot sell in the EU. Therefore, it is not only relevant for EU-based companies, it is also very relevant for US-based companies or companies in Asia that, want to sell their ISCO products also in Europe.

So, as, Andrew already alluded to, like, the majority of US companies will likely be manufacturers of certain in-scope products. They will want to keep selling them in the EU. That's why all those products need to comply with the requirements of the CRA. What those are, we will touch upon in a second. But that's kind of, like, the major aspect. If you're producing and manufacturing, in-score products, then this is very relevant for you.

It can also be very relevant if you have, like, an used subsidiary that is placing those products, on the market, then they may qualify, in the legal sense, as the manufacturer, so to speak, like, as quasi-manufacturer of those, or as, or as an importer, or a distributor. So, they maybe also be caught by, by the requirements, either directly or as your representative, in that sense, in the EU. And then you have all those, indirect implications.

First of all, if you are supplying… if you're not putting, placing those products on the market yourself, but you have, like, a specific importer or distributor in the EU, then they kind of… they switch roles with you, and then, from a legal perspective, they are regarded as the quasi-manufacturer for the purposes of the CRA, and they will need to comply with all those requirements. They cannot do it because they're not the de facto manufacturer, so they will Basically, on a contractual basis, ask of you to kind of do all this stuff. do all those requirements, provide them with documentation, provide them with security updates, etc, etc.

So, if you're caught in one of those scenarios, then you can have, like, a very indirect, applicability of the CRA requirements to you. The same applies if you're a supplier of components, that are then built into CRA-relevant products, by, for instance, EU companies that are then then those products will be subject to the CRA, then you might be caught also indirectly as a supplier, because, as we will see in a second. InScope products must be evaluated also regarding all third-party components that are built into those… into those products.

So, you need to, evaluate what is the… the risk exposure of these product… of these components, which risks arise by using certain components from certain manufacturers, how can those be mitigated, etc, etc. So. You will get, as a supplier of those components, you will get, a lot of, lot more questions from, from your customers regarding, what is the security, what are the security aspects of this component? Are you able to provide security updates regarding this component for, let's see, the next 5 years, etc, etc.

So, there will be a lot more, due diligence and scrutiny on suppliers of components, as well as more, contractual requirements in that regard. So, to… to summarize. Very important, if you make… if you're doing business in the EU, the CRA is very relevant for you, even if you're not established in the EU.

Andrew Lewis: Right. Alright, so we've talked about scope, we've talked about applicability, and so now we're gonna launch into what does it actually require you to do? And so these 6 boxes, it requires you to do a lot, and these six boxes, these 6 categories, don't cover everything, but they cover a lot of it, and I think it's helpful to think about it in these sort of, categorical ways. And we're gonna talk about each one of these in depth, but wanted to give you a sort of overall view of what CRA requires. And what we're gonna start with is the cybersecurity risk assessment for the product, because that sort of is the backbone of everything.

It requires… Manufacturers to document the product cybersecurity risks and use that assessment throughout the life cycle of the product. Including with relation to the vulnerability handling of the product, which we'll talk about, and the secure product design. So, how you then incorporate cybersecurity aspects into the product is really dictated by that risk assessment that you conduct. at the beginning, and throughout the lifecycle of the product. That requirement is… Requires updates to it with, sort of, changes to the product itself.

The other aspects here are the, the conformity, so how you're actually preparing your technical documentation, submitting that, whether or not you're going through a self-assessment or bringing in a third party to do that, and how you're actually affixing, you know, the marking to your product in the EU market. The incident and exploitation reporting we'll also discuss, which is the one aspect of the CRA that is currently live. And then, of course, the post-market responsibilities.

It's sort of what you have to do once you launch the product, and maintain, throughout the life cycle of the product. how you determine what the life cycle is, and what sort of the minimums are for the lifecycle. So that's kind of our roadmap… roadmap for the rest of the day, and we'll… we will obviously, of course, talk about penalties, and enforcement actions at the end of this, too, as well, because that's everyone's favorite topic. But with that, let's get started on this. So the… like I mentioned, the risk assessment really is the backbone of all of this.

So it is a required piece of this and dictates a lot of what happens with the product assessment. So… Manufacturers are required to assess the cybersecurity risks associated with the product and include the assessment in the technical documentation. the assessment itself needs to take into account, needs to be taken into account during the planning, design, development, production, delivery, and maintenance. So all of those aspects of the product, need to be sort of driven by this assessment. The… assessment also needs to evaluate the relevant risks in light of the product's intended purpose and reasonably foreseeable use.

So this one's important, too, is companies need to think about both what they're intending the product to be used for, and any reasonably foreseeable use. So, that may be something that, you know, you're not necessarily intending, but it could be part of what it's used for, and you need to take that into consideration when you're conducting the risk assessment. So things that… You might not think about when originally designing the product, if it comes to light that, you know, your customers may be using it for, may dictate a sort of reassessment of that and re-evaluation of the security features that need to be in the product because of that.

And that brings us to this updating portion of it. So. as a manufacturer, I need to revisit the assessment as… product functionality, threats, vulnerabilities, dependencies, or intended use evolve. And in, you know, the age of AI, this is, I think, an important one because things are changing so rapidly that that re-evaluation may occur quicker than I think we were originally intending when this law was written. And again, so just something to think about when building CRA compliance is… how do we revisit this?

And how do we revisit this quickly once we launch this product, because after 3 months, we learn of XYZ vulnerabilities, or this is being integrated for certain reasons, or people are just using this in a way that we didn't think that they would? Kind of pointing back to your risk assessment. Before any of those things happened, and saying, well, we weren't thinking about that at the time, is not going to be enough. Because there is this… Expectation and requirement to continually update the risk assessment. And just as a basic sort of, practice point here.

Retention, and evidence of both the original risk assessment The policies and procedures designed to address these updates, and the updates themselves, should all be well documented and maintained throughout the lifecycle, because these are going to be the first, sort of, pieces and artifacts that are requested by, regulators. We'll get to this in a second, but just a, you know, sort of overarching reminder that the risk assessment is the backstone of most of this. And drives a lot of the other requirements. So those technical… things that you have to implement are really driven by what are the risks associated with this product.

So, the level of… protection for confidentiality, availability, integrity of data is driven by it. How you need to minimize the processing of the data is driven by it, and a lot of that goes into what kind of data are we talking about that this has access to, and the requirement to sort of minimize the negative impact of the product on other devices. all ties back to that risk assessment, and what that means is, is this thing that we are launching going to have the ability to interact significantly with other software? And if so, then stronger security needs to be put in place.

Okay, and I will… so the secure product design, I'll turn that over to you, Tobias.

Tobias Rothkegel: Thank you. Yeah, and as a next step, as Andrew said, when you have this risk assessment basically carried out per each product, or product range, or product family, whatever, however you want to… you can structure it, you need to translate this into a secure product design. This is kind of like the gist of the CRA, where you can say you're making conscious risk assessment about product, and then you can translate it into how can we make this as secure as possible.

This is also risk-based, so it doesn't have to be In totality, as secure as possible, but just, you know, appropriate to the risks involved with the functionalities, with the criticality of the product, etc, etc. as we will see later, there are also different classes of products within the CRA, so kind of as a rule of thumb, is the more, critical in class my product is, the more attention I have to kind of pay towards any of those requirements, right?

So, And, if you're interested in, looking at the requirements in more detail, NX1 to the CRA is very important because it basically lists all the requirements that a manufacturer needs to take into account when designing. a product, and for the most part, we'll also need to make sure that this is actually possible. So, if you look at the requirements, for instance, it is, it's, product shall not be made available on the market with known exploitable vulnerabilities, so if you have knowledge about vulnerabilities in the product, you cannot basically put it on the market.

Then you need to have a secure by default configuration, which basically means, by default settings, it has to be configured as secure as possible, like, all ports have to be deactivated, for instance, or if it, if it allows MFA, MFA must be… must be enabled, etc, etc. So, if it reaches, when it reaches the customer, it has to be, configured as secure as possible. This is kind of, like, one of the… one of the major, cornerstones of those, of those requirements in Annex 1. Then you need to have, you need to basically implement measures that, reduce the potential attack surface of the product.

So, it, needs to be, as I said, as little pods as possible, as little accessibility as possible, as little interference as possible. It has to have… appropriate, authentication methodology, et cetera, et cetera. So, it, must be, designed and then manufactured in a way that, reduces the attack surface as much as possible. This also means that, if you need to design the product in a way where, when vulnerabilities occur, so I think it's also very important to understand. the CRA does not require that a product may not have any vulnerabilities.

It is basically, granted that every product that will come to the market will have certain vulnerabilities. It only says you need to Exclude any known vulnerabilities, and you need to design the product in a way that most vulnerabilities that occur over time, or get known over time, can be addressed by security updates. So, We had an instruction, for instance, by a client where they said, well, we have this vulnerability. but it cannot be addressed on a software level, it is a hardware thing. So, basically, it's not… then, it wouldn't really, comply with the CRA.

Like, looking forward, the CRA requirements are not applicable yet, but, looking ahead. That would be something where, in the design phase, you would need to figure out, well, we need to reduce, The vulnerabilities that may occur through hardware components within the product, so you need to design it in a way that it can be addressed through updates as much as possible. Then you have to put in, appropriate protection for any data and functions within the product. This is also, as Andrew said, all very risk-based driven.

It depends on the criticality and sensibility of the data. that, is, stored within on the device, as well as the… as the overall functioning. So, if you have, like, a critical component in a, in a power generation facility, then you would have higher, of course, a higher degree of protection, whereas if you have, like, a home appliance, you can… you can slack here and there, so to speak. So, it all… all… all… it kind of also goes down to, what the… What the involved risk is.

And then you need to, last but not least, you need to make sure that the product overall is resilient, and if something occurs, then it is easy to contain. usually by, recovery procedures or by the possibility to deploy, security updates and, fixes and patches in a timely manner, et cetera, et cetera. So, what that means in the end is that you need to, have a look at those requirements in Annex 1, And then translate this into, release criteria for each product, and version where I kind of say every product and version needs to go through certain gates and pass certain tests before it can actually hit the market.

And this is all very… Individual, based on the product.

Andrew Lewis: So, as Tobias has been referencing, this Annex 1 is a really good read. It's only a couple pages, actually, so in typical EU fashion, there's a very long, document, and then it sort of points to, the annex and says what you actually need to do. And so, in the text of the law, it has the risk assessment requirement, and then it points to the annex, which is divided into… Annex 1 is divided into two parts. what Tobias just described, the sort of secure design portion of it that is, tied to the risk assessment, and then this vulnerability handling portion, that I am going to talk about right now.

So, the risk assessment. if we haven't already made it clear, is very important, and should be sort of the backbone of what you're doing here, because all of the decisions that you make on how to handle both your vulnerabilities and the design of your product, need to sort of be justified by the risk assessment. So getting that right Early is really important.

Okay, let's jump into it with the vulnerability handling, because this is one of the more unique aspects of the law, and part of what… It is designed to assist with, is this idea of… patching things quickly and updating customers, and making sure that you have a good sense of what is in your product. So the first sort of requirement here is to establish processes to identify and record vulnerabilities and product components. And there is a requirement to have a software bill of materials here, so a sort of documented roadmap of the contents of your product. That needs to be in a commonly used machine-readable format.

We're seeing some of this on the US side of things as well. There is a push to have this… documentation and this sort of supply chain security shored up, and this is… I think we're gonna be seeing more of this, so… taking the CRA as an opportunity to go through these processes at your company, I think, is a really good, way to address The trajectory of what we're seeing here. And the idea here is, back to our point about open source, right? You may be taking an open source product and putting it into your software.

That then triggers, if you're providing this software or this product, through commercial means in Europe that is going to trigger the CRA, which means you are going to need to do this sort of analysis on that open source piece. And so knowing that it's there to begin with is going to be a really important function, and why they're sort of directing this identification early on. You also need to be able to receive and disclose vulnerabilities, so you need to… this is sort of… almost nitpicky procedural practice stuff, to put in here, but I think important, because not all companies are going to have These types of… you know.

Public policies or mechanisms, established already. And this may be a new thing that you have to implement, but you need to have sort of a contact point. This is part of the documentation for the submission to the regulator, is that you have to have this contact point, and a policy for reporting, and how you then handle the vulnerabilities that are reported to you. So, not the biggest lift, but just something new that you may not have, already established. Identified vulnerabilities need to be addressed, without delay. And that requires both supplying an update and communicating any relevant corrective information to users.

That without delay, obviously, is used often in European regulations. We don't have a real hard definition of it, but again, back to sort of the risk assessment here, if… sending out something before you're able to fix it is going to cause more harm than good. Some delay may be useful, but the idea is that, you are addressing these quickly, and communicating what needs to be done from the customer perspective. It's another one that you may need, to establish those communication channels, if you don't have them already.

You know, if you are offering a SaaS product that comes under the CRA, you may be able to communicate with your customers more actively than if you are, you know, selling products off the shelf. So thinking about how you do that communication is another just sort of step to, to, work through. The… Updates need to also be disseminated securely, and separately from functionality updates when applicable. The support period is also just part of this overall CRA identification, is that you need to have a declared support period that reflects the expected product use and, Is in line with… how you expect customers to use your product.

That's generally… 5 years? But you can sort of make a case that your product will not be used for longer than 5 years. Again, you need to take into consideration sort of unexpected use cases. So I think you'll need to defend If saying anything less than the 5-year kind of assumed threshold within the CRA. And sort of just some tips and tricks on the vulnerability handling.

So, obviously publish the monitored security contact, because that's a requirement, and have that disclosure policy also available. do the SBOM early, update often. preserve all the links between your components, suppliers, products, and, versions, which I think is important, because as you update the version, things may fall off, reassessment, may need to… may need to occur. Security updates themselves must remain… so this is not the lifecycle, but the actual updates themselves have to remain available for a minimum of 10 years, or the remainder of the support period, whichever is longer.

And security updates must be made free of charge, unless there is… there's an exception for this for, like, tailor-made products, but generally speaking, you can't charge for the security updates. Okay. Tobias, do you want to talk about our conformity and market access?

Tobias Rothkegel: Sure. So this is… this slide is basically to… intended to give you an overview of, like, the very broad steps you need to ensure and take, when you want to bring a product, an InScope product, to the market. So, it obviously starts with classifying the product, which. The first step is, is the CRA applicable to the product in the first place?

Which… A lot of times it's very easy, because if I have, like, an IoT device, it's very apparent that this is covered by the CRA, but it can be way more complicated than that, particularly if you have very complicated machines that consist of several different components that are then also marketed separately, etc, etc, when, you can also have products that interact with, other, IT systems, which are already in place with the customer, you may have, in a sense, non-digital, products that have some sort of, communication, component then attached to them, so you can have, like, connected devices, you can have tailor-made products, etc, etc.

So the assessment whether a, a product subject to a CRA can also be very complicated. The more complicated the products are, usually it is more complicated to actually assess whether the CRA applies to them in the first place. But once you have established that the CRA applies. then, as I already, alluded to before, is that there are certain different classes of products within the ZRA.

You have basically, let's call them the default class or the normal class, then you have the, important products, there are also two classes then, Class 1 and Class 2, and which are, for instance, password managers, boot managers, operating system, routers, modems and switches, microprocessors, etc, etc. Firewalls as well. So there are certain, an exhaustive list, so to speak, of, Of products that fall within the, important products category.

And then there are also, critical products, which is kind of like the highest The highest class, which are, for instance, hardware device security, boxes, smart card readers, smart meter gateways, and, Those classes can be extended by, implementing regulations. So, if the Commission at some point decides, well, we need to expand those classes, they can do that, without actually touching the CRA, but by, enacting a, an implementing regulation, so, there might be still needs to be some sort of, you know, monitoring in that regard.

What's also, also on the horizon, at least that's the intention of the Commission, is to have specific, additional requirements for CRA-relevant products that are used in critical infrastructures, for instance. So, there might be also additional, requirements coming into force, after, end of next year, when the CRA kicks in.

So once you have kind of classified the product, or the product range, or product family, then the, the classes of the products, are, as I already said, relevant for the… for the risk assessment, where I say the more critical my product is, the more Usually, the more risk is involved, the more robust of security measures I need to observe and implement. But also, there are certain different certification methods baked into this ZRA. So, for instance, for the default class, so products with digital elements that are neither an important product nor a critical product, the… you can use self-certification, so to speak.

There's nothing to… basically no structure to follow. You could do it on your… on your own accord. You need to just, Document your, your process and your decision, but you can do it on your own, so it's a self-certification. But then, the more you go down, kind of like the, or up the criticality, if you want to see it in that way, for instance, for important, products, you need to follow certain, certification schemes, certification protocols, to, to fulfill the requirements regarding important products. And, for instance, for critical products, you need a, certification by the… by the INISA in between.

So, for certain, also important products, there's also a… also a distinction within the law that if your important product fulfills certain, requirements, if it has a specific, risk exposure, for instance, then you also might need a, third-party certification from a accredited certification body.

So, that's something also to factor in on a times… on a timeline, where you kind of say, well, I do the… my product portfolio analysis and see, well. 90% of the stuff I can do within self-certification, that's good, but for certain products, I might need third-party certification, so there's something where you need to factor in some time to actually obtain this third-party certification, because those, certification bodies will be very busy by the end of next year, I can imagine.

Then before you go to market, you need to compile all the, technical documentation, as Andrew already touched on, where you kind of say you need to… you need the SBOM, you need, manuals, etc, etc. You need the risk assessment, you need the controls. that's usually something that needs to be provided alongside with the product to a certain extent. Then you need to have your declaration of conformity ready, which is where there's a template, you need to just basically fill it out.

And you need to affix the CE marking on the product, and of course, you need to preserve all of this, document all of this, and Maybe just a takeaway is that it is very useful to start this journey rather early, and to get, like, a good grip on my product portfolio, which products are in scope, etc, etc, because down the stretch, you not only need to look at the product in itself, you need to look at all the components, you need to make risk assessment for all those components. Potentially, you might also need to, Re-evaluate any supplier engagements, etc, etc, so this is something that might take, or will take, a lot of time.

Andrew Lewis: And thank you, everyone, for questions. By the way, we will address those at the end of the presentation, so please, keep sending those in. Daniela is cataloging them for us, and we will, address them when we get, to the end of this. So, but thank you. I have some really good ones so far, so appreciate people asking questions. Okay… On to everyone's favorite, incident and exploitation reporting. As we've mentioned earlier, this is the one part of the CRA that's already live. So this went live September 11th of this year. And so, for those that have products in the European market, you need to be making these notifications, as of right now.

What's interesting about the CRA is there really are two notification requirements built into the law. And that's for severe incidents and then actively exploited vulnerabilities. The severe incident we'll talk about in a second, that sort of tracks other notification requirements. It's very similar to other things that we've seen or are going to see in Europe. The actively exploited vulnerability, though, is a bit different, and I think it's the one that is, causing both confusion and consternation, from companies trying to implement this. You know, importantly. The definition requires evidence and reliable evidence of an active exploitation.

So, something actually needs to have happened. So this is… this is not just a, we think this… this is, you know, after a reasonable investigation, we've seen evidence of the vulnerability being exploited. What's different, though, is that it's not really tied to an impact, so I think it's somewhat both broader and narrower, right? It has to be tied to a vulnerability, but it also isn't tied necessarily to… To PII, to customer content, like we normally see. So, just sort of getting through the door is enough. For this notification provision to be triggered.

And importantly, it's one of the… unless you're subject to something like FedRAMP that has a 1-hour notification time window, the 24-hour notification early warning submission is quick. So… Having processes in place that can identify the vulnerability, kind of get that, initial… information back to you, and then actually being able to physically submit it within 24 hours may be something new to you. Again, unless you have sort of a robust reporting obligation under something like, you know, CMMC, FedRAMP, that requires sort of a shorter timeframe. The 24-hour might be the shortest time frame that you have.

There is also this, and both the actively exploited vulnerability and severe incident notifications, they are separate notification requirements, but they generally follow the same notification, procedures. So you have that 24-hour for both of them, you then have the 72-hour sort of update, if you will, and that's gonna… there's certain information that needs to be submitted throughout that. And then you also have this final report, and… Affected user, update as well.

So… I think the takeaway here is… If this is your shortest, you know, this is going to dictate a lot, especially if this is your shortest timeframe, because what you're going to want to do during this is also consider the other reporting requirements that you may face based on whatever it is that you're seeing. So… think about CRA, think about NIST2, think about GDPR, DORA, any of your US laws, while you're doing this. Test the actual reporting platform, because it is, and I've heard that it has some hiccups.

They launched an actual… reporting, like, single reporting platform in order to make these reports, so test that out, make sure that the person that, you know, that you have a team that can actually submit this, in a timely manner, and sort of have those templates drafted so that you can get this done in a quick, in a quick manner. Okay, and post-market responsibility. So, we touched on this a little bit, so I'll go through this a little quicker, because we've kind of been talking about this throughout. The important part of this is just that you don't just do the risk assessment, put the product on the market, and say, good luck.

There are requirements to Maintain vulnerability reporting, bring non-compliant products back into conformity, when necessary, so if you learn that something is out of compliance, you need to take corrective action. You need to, make sure that products are receiving security updates. That's a big part of the lifecycle requirement here, and as we said earlier, it's sort of a five-year… the law itself says 5 years, unless you can demonstrate, and this would be part of that risk assessment, that the product itself is not going to be around for that long.

But again, you need to make… go through those… motions and make sure that you're, identifying that. And there was a question earlier in the chat about the 10-year requirement. The 10-year requirement is not for ongoing security updates. it's if you supply a security update for something, that update itself needs to be made available for 10 years. So, the product may go away, but you want to make sure that your customers can update it for a 10-year period. That's… that's where the 5- and 10-year, sort of, timeframes come into play.

And as we talked about. a couple of times is you need to continuously identify the changes in these risks, and, especially where there's either modifications, new functionality added to the product, ownership changes could affect the way in which the product is evaluated. All of these things are going to trigger sort of a new assessment, and so you need to make sure that you have, again, those processes in place in order to be able to Identify when those things are triggered, and go back to the drawing board. Okay. And Tobias, do you want to just kind of briefly talk about, like, where this sits in our broader EU framework?

Tobias Rothkegel: Yeah, sure, just to… just to… that you get, like, a better understanding what the… what the landscape is around cybersecurity regulation right now in the US. So, the CRA, what we've been talking about now, it's very product-focused, like, products with digital elements, looks at manufacturers, predominantly. Then you have NIST 2. NIST 2 is limited to the, so to speak, critical infrastructures, so they're looking at companies and their activities, so not products specifically, but they're looking at the activities of a company. Then, of course, you have GDPR, which everybody knows, processing of personal data. Then you have DORA.

DORA is basically the NIST 2 for the financial sector. So if you're a financial company or an insurance company, you will be subject to DORA, and you're basically exempt from NIST 2 then. And the AI Act, we've, put it in here. It's, I mean, it, it applies to AI systems. But AI systems may also be incorporated into CRA-relevant products. So, what the AI access, it also has cybersecurity requirements, specifically for high-risk AI systems, but then it also says, if your AI system is also subject to the CRA, it's sufficient to comply with the CRA requirements. To also, fulfill the AIC requirements in that regard. For cybersecurity.

Andrew Lewis: Perfect. And then, everyone's favorite topic, penalties and enforcement exposure to bias, if you would run through this for us.

Tobias Rothkegel: Sure, yeah, so we see GDPR level, fines here, and, so we have 3 different brackets, as you see in the top portion of the slide. What is interesting is that the highest brackets, so the $50 million over the 2.5, percent of annual turnover. Both Article 13 and 14 sit in that high bracket, so which are the manufacturer requirements, which we've been talking to, as well as the vulnerability handling as well as notification requirements. So, basically, the two major, obligations sit in the highest bracket of the, Of the fines. But I think the fines aren't really the most, most crucial thing to think about.

I feel that what will be more, substantial is That the regulators also have the power to pull products from the market. So, if there's a formal discrepancy or even a material violation of the CR requirements, they can basically pull products from the market. This is what the authorities may do, but we also need to consider that this will be a antitrust, unfair competition law aspect as well. So.

I wouldn't be, surprised if competitors start going after each other, when one has basically retrofitted all these requirements and fulfills all the CRA requirements, then sees a competitor that doesn't, maybe also maybe from for instance, from another country, etc. So there might be, as well, civil law action, between, competitors themselves. And also, they may also act on regulators to kind of say, well, we have seeing certain… certain vulnerabilities and potential violations with products from company XYZ, so this may also also happened, so I feel that the, the real risk is rather, civil action rather than, regulatory exposure.

This leaves my gut feeling.

Andrew Lewis: Okay, we have one minute left, so I think we only have time for one question, but I did get a good one in the chat that I will, relay here, is… This and a lot of other requirements seem duplicative. Other laws like GDPR, the CCPA's new cybersecurity audit requirements, can we leverage those assessments in putting together the requirements under the CRA? And I think, yes, I think that gets kind of the heart of all of this, is that what we're seeing is a lot of new Regulations, requirements that are focused on the security, Of products being… entered into certain markets. And especially in the situation where you are doing a self-assessment.

A lot of these things are asking for the same thing. They're asking for this risk assessment, they're asking for there to be security by default, and I think you can accomplish that from a holistic design perspective. You just need to think about it sort of from the beginning, and so that risk assessment, I think, really will play into how you're addressing all of these laws, because you're not going to want to do every one of these things 15 times. You're going to want to do it once, and make sure that it maps across all of your requirements. And so having that, that sort of discussion early is important.

And with that, we're right at 11 o'clock, so thank you all for being here. Tobias, thank you so much for staying up late for this one. I'm gonna leave up our contact information. Please feel free to reach out for questions, comments. We really appreciate everyone attending today.

Related Articles

DWT logo
©1996-2026 Davis Wright Tremaine LLP. ALL RIGHTS RESERVED. Attorney Advertising. Not intended as legal advice. Prior results do not guarantee a similar outcome.
Media Kit Affiliations Legal notices
Privacy policy Employees DWT Collaborate EEO
SUBSCRIBE
©1996-2026 Davis Wright Tremaine LLP. ALL RIGHTS RESERVED. Attorney Advertising. Not intended as legal advice. Prior results do not guarantee a similar outcome.
Close
Close

CAUTION - Before you proceed, please note: By clicking "accept" you agree that our review of the information contained in your e-mail and any attachments will not create an attorney-client relationship, and will not prevent any lawyer in our firm from representing a party in any matter where that information is relevant, even if you submitted the information in good faith to retain us.