The Linux Foundation Stewardship Framework FAQ
About this FAQ
The answers below describe how the EU Cyber Resilience Act (Regulation (EU) 2024/2847, the “CRA” and the Linux Foundation’s stewardship framework work, drawing on the CRA text, the European Commission’s CRA guidance (C(2026) 5252, July 2026) and implementation FAQ, and ENISA’s Single Reporting Platform (SRP) materials. The information below is general information for the LF community. It is not legal advice and does not address the circumstances of any particular project, maintainer, employer, or company. LF Legal represents the Linux Foundation and its project-hosting entities; it does not represent individual contributors, projects, or member organizations. If you need advice about your own situation, please consult your own counsel. Where the CRA or official guidance is unclear, we try to say so.
Who carries the CRA reporting obligations, and what should a maintainer do if an escalation to the LF CRA Steward is late?
The CRA places its reporting obligations on legal persons: manufacturers (Article 14) and open source software stewards (Article 24(3)), not individual maintainers. Under the LF framework, the LF CRA Steward is the legal entity that files reports through the EU Single Reporting Platform. Generally, a project community or an individual maintainer is not the party that files. The exception is a project that has elected, by documented agreement with LF, to operate as its own steward and has registered its own Assigned Representatives on the SRP.
The CRA’s 24-hour and 72-hour deadlines are measured from when the steward becomes aware of a reportable event (see the next question). The CRA also states that it does not apply to individuals who contribute source code to free and open source software that is not under their responsibility (Recital 18; Commission “CRA - Open source” page. Separately, Article 64(10)(b) provides that the CRA’s administrative fines do not apply to any infringement of the Regulation by open source software stewards.
An LF project should escalate to the LF CRA Steward as soon as it has credible information of a reportable event. The LF CRA Steward makes and records the awareness assessment and files it with the SRP. If an escalation reaches the steward later than it should have, the right response is still to escalate immediately and tell the steward what you knew and when. A late escalation may make the steward’s job harder; but it never changes the first priority, which is fixing the problem.
When does the CRA’s 24-hour early-warning period begin? Is it when an email is received, when it is read, or when exploitation is confirmed?
The CRA measures the 24-hour early warning and the 72-hour notification from the moment the reporting entity “becomes aware” of an actively exploited vulnerability or a severe incident (Article 14(2) and (4), applied to stewards by Article 24(3)). The final report runs from a different point: 14 days after a corrective or mitigating measure is available for an actively exploited vulnerability, or one month after the 72-hour notification for a severe incident. The Regulation does not define “aware”. The Commission’s guidance (points 209-221) explains that an entity that detects a suspicious event, or receives a report from a researcher, customer, authority or the media, should assess it immediately, and becomes aware when, after that initial assessment, it has a reasonable degree of certainty that a vulnerability in the product is being actively exploited or that a severe incident has compromised the product’s security. The Commission’s FAQ subchapter 5.1 explains how a manufacturer may become aware of such a reportable event, for example through reports from customers, partners, security researchers or cybersecurity authorities, threat intelligence reports, or its own monitoring, scanning or telemetry activities.
“Actively exploited” is defined by reliable evidence that a malicious actor has exploited the vulnerability in a system without the owner’s permission (Article 3(42)). So, an unread or unverified email does not by itself start the clock, and neither does a suspicion no one has examined; but the guidance also expects the initial assessment to be done promptly, particularly where the risk may be significant. A published CVE identifier, on its own, is not evidence of active exploitation.
For LF projects, the practical rule is simpler than the legal text: do not wait for proof. Escalate to the steward as soon as you have credible indications of exploitation, or of a compromise of your build, release, or signing infrastructure, with whatever information you have. The steward will perform and document the awareness assessment; your escalation is an input to that assessment, not an admission of anything.
The CRA’s implementation is still in transition. Manufacturers have been subject to Article 14 reporting since 11 September 2026, while the Commission and ENISA have confirmed that steward reporting obligations under Article 24(3) apply from 11 December 2027 (Article 71(2); Commission FAQ 5.5; ENISA SRP FAQ). The LF stewards are putting their processes in place now so the escalation path is familiar well before then.
What does “ultimately intended for commercial activities” refer to, and how should a project think about it?
This question looks at the intended use and ecosystem of the software, not at who the end user is. The CRA gives integration into commercial services or monetized products as the typical example, but the phrase is broader than that. What matters is whether the software is expected to end up in someone’s downstream commercial activity, even if the project publishes it for free. Software intended for companies or organizations to use is generally covered. So is software that people use directly, such as a desktop application, a browser, an operating system distribution, or a mobile app, where companies are expected to ship it in their products, bundle it with devices or services, offer it as a paid or supported edition, or otherwise build commercial offerings around it. Conversely, software published only for non-commercial use, such as a hobby project, research prototype, or teaching tool, is not. While the Linux kernel started off as “just a hobby, won’t be big and professional like gnu”, it has changed over time to be intended for commercial activities.
Developing free and open source software within a not-for-profit organization is not a commercial activity provided earnings after costs go to not-for-profit objectives, and donations, sponsorships, or grants do not change that (Recital 18). Nor does a company’s decision to use your software in its product make your project a manufacturer; that company is the manufacturer of its own product.
Projects that develop and publish best practices guides, standards, data sets, or specifications are not publishing software intended for commercial activity and therefore not required to report to the SRP. If an open specification project also publishes an implementation, the implementation could be intended for commercial activity.
Within the LF framework, this question only determines whether a project needs a stewardship arrangement. If the answer is unclear, ask LF CRA Steward through your program manager. The LF CRA Steward is responsible for determining the scope of its own stewardship and will make the final call on which projects it covers.
How do we contact the LF CRA Steward?
Email steward@linuxfoundation.org. Use it to escalate an actively exploited vulnerability or a severe incident, and for questions about the stewardship framework (which entity is your steward, classification, or an election to self-stewardship). It is not a general vulnerability intake: ordinary vulnerability reports should go through your project’s own security reporting mechanism. When escalating, send what you have, in the order you have it: a description of the issue, affected versions, any evidence of exploitation, reproduction steps, and known mitigations. Do not hold an escalation to complete it. For help identifying your steward or updating contacts, you can also use support@linuxfoundation.org.
Does the CRA require a particular open source license, or change our project’s license?
No. The CRA does not mandate any license, and it does not alter the terms of the license your project already uses, including its warranty and liability disclaimers. What the license does affect is whether software counts as “free and open-source software” for CRA purposes. Article 3(48) defines that as software whose source code is openly shared and which is made available under a free and open source license that provides for all rights to make it freely accessible, usable, modifiable, and redistributable; the Commission’s guidance (point 44) confirms that both conditions must be met. Software released under a license that withholds some of those rights, for example a non-commercial or “source-available” license, may not qualify as FOSS under the CRA, and the steward regime is written for FOSS. LF software projects use OSI-approved licenses and publish their source, so this is not an issue for them. No license change is needed or recommended for LF software projects in connection with the CRA.
If a company uses our software in a commercial product, does the CRA apply to our project?
Not on that basis alone. The CRA regulates manufacturers, who place products on the market, and the legal persons who act as open source software stewards. A company that integrates your software into its product is the manufacturer of that product. It is responsible for its product’s conformity, must exercise due diligence when integrating open source components (Article 13(5)), and, if it identifies a vulnerability in your component, is required to report it to the person or entity maintaining the component and to share any fix it develops (Article 13(6)). Those are the manufacturer’s obligations, not the project’s.
An open source project’s own position depends on who supports it. Providing free and open source software that is not monetised is not a commercial activity (Recital 18), so a project that simply publishes software is not placing it on the market. Where a legal person systematically supports the project on a sustained basis (for example, by hosting, governing, or providing development infrastructure for it) and the software is intended for commercial activities, that legal person may be an open source software steward with the Article 24 obligations. For LF-hosted projects, that legal person is the LF project-hosting entity, which is why the LF entities are acting as stewards; the project community and individual contributors are not the regulated party in that arrangement.
You should expect that downstream manufacturers may ask more of you, for example, security contacts, SBOMs, attestations, and timely advisories. Those are not CRA obligations on the project, but providing these features could be best practices to support your manufacturer ecosystem. Additionally, establishing clear intake channels enables your project to capture helpful enterprise contributions (like PRs and fixes).
How do the CRA’s obligations for open source software stewards differ from those for manufacturers?
Manufacturers (Article 13 and Annexes I and VII) must design and build products to meet the essential cybersecurity requirements, carry out conformity assessment, affix the CE marking, prepare technical documentation including an SBOM, handle vulnerabilities throughout a support period, exercise due diligence over third-party components, and report actively exploited vulnerabilities and severe incidents under Article 14. They can be subject to administrative fines.
Stewards (Article 24) have what the CRA calls a “light-touch and tailor-made” regime (Recital 19). A steward must put in place and document, in a verifiable manner, a cybersecurity policy that fosters secure development and effective vulnerability handling and encourages voluntary reporting; cooperate with market surveillance authorities and provide that documentation on reasoned request; and report actively exploited vulnerabilities and severe incidents only to the extent set out in Article 24(3): actively exploited vulnerabilities to the extent the steward is involved in developing the software, and severe incidents where they affect network and information systems the steward provides for development.
Stewards are not subject to essential requirements, conformity assessment, CE marking, SBOM or technical-documentation obligations, or administrative fines (Article 64(10)(b)). A market surveillance authority that finds a steward non-compliant may require corrective action (Article 52(3)). A steward is by definition “other than a manufacturer” (Article 3(14)).
What does the CRA say about fines for open source projects and stewards?
The CRA’s administrative fines (Article 64) are addressed to economic operators, principally manufacturers, importers, and distributors, and can be substantial. Article 64(10)(b) provides, by way of derogation, that those fines “shall not apply to … any infringement of this Regulation by open-source software stewards”; the Commission’s own CRA pages say the same. The CRA does not impose fines on open source projects as such, because a project community is not an economic operator under the Regulation, and the CRA states that it does not apply to individuals contributing code to free and open source software not under their responsibility. The remedy against a non-compliant steward is corrective action required by a market surveillance authority (Article 52(3)).
This describes the CRA only; other laws, such as national product liability or contract law, are outside this FAQ. Some have their own treatment of open source. For example, the revised EU Product Liability Directive (Directive (EU) 2024/2853, Article 2(2)) excludes free and open source software developed or supplied outside the course of a commercial activity, and places responsibility for an integrated open source component on the manufacturer of the product that integrates it (Recitals 14 and 15). Note also that a company that places a product containing open source software on the market is a manufacturer under the CRA and is subject to its fines as such, regardless of where the code came from.
Do we need to escalate every vulnerability or CVE to the LF CRA Steward?
No, and please don’t route routine vulnerability reports to the steward.
Only two kinds of events need to reach the steward: an actively exploited vulnerability, meaning there is reliable evidence that a malicious actor has exploited it in a system without the owner’s permission (Article 3(42)); and a severe incident affecting the project’s development infrastructure, such as a compromise of your build, release, or signing systems, or anything else that has led or could lead to malicious code in a release (Article 14(5)).
Ordinary vulnerabilities, including those with CVE identifiers and those with a public proof of concept but no evidence of real-world exploitation, go through your normal process: fix, release, and disclose as you do today. Voluntary reporting to the authorities under Article 15 exists for anything else; the steward can discuss that with you if you ever want to use it. If you are unsure whether an event meets either threshold, escalate it to the steward if you have credible information suggesting active exploitation or a severe incident. You do not need to establish conclusively that the event is reportable before escalating it. Routine vulnerability reports, however, should continue to follow your normal process.
Who communicates with users about a vulnerability or incident: the project or the steward?
The project does, through its established disclosure process, and that does not change. Advisories, embargoes, coordinated disclosure with downstream distributors, release notes, and mailing-list announcements remain the project’s responsibility, run the way the project runs them today. The LF CRA Steward’s role is the regulatory interface: filing with the EU Single Reporting Platform and dealing with CSIRTs and market surveillance authorities.
Article 14(8) requires a reporting entity to inform impacted users, and where appropriate all users, of an actively exploited vulnerability or severe incident and of any mitigations; for stewards that applies to the extent described in Article 24(3). The Commission’s guidance (point 82) recognizes that stewards typically do not have direct relationships with users, so the steward will satisfy this through the project’s normal public channels and will coordinate timing with you so that regulatory reporting does not break an embargo. The 72-hour notification lets the reporting entity indicate how sensitive it considers the information (Article 14(2)(b)), and the steward will use that. Downstream manufacturers who ship your software have their own Article 14(8) duty to inform their customers; the project is not expected to notify a manufacturer’s customers, LF members, or anyone else on their behalf.
What other best practices does OpenSSF recommend?
While not specifically required, as part of fostering secure development for its projects, stewards should help their projects adopting baseline security hygiene (such as OpenSSF best practices and tools). Projects can implement and document these voluntarily without expanding project or steward liability. You can learn more here: https://policy.openssf.org/CRA/maintainers.html. Even without statutory penalties, unaddressed security risks can damage project trust and its adoption. Focus should remain on sustainable project health and good engineering and security practices.
Where can our company learn more about the CRA?
LF Legal acts for the Linux Foundation and its project-hosting entities and cannot advise member companies, maintainers, or downstream users on their own obligations under the CRA; for that, please consult your own legal counsel. Good general starting points are the OpenSSF CRA resources at https://openssf.org/public-policy/eu-cyber-resilience-act/ and https://policy.openssf.org/, and the Commission’s CRA pages at https://digital-strategy.ec.europa.eu/en/policies/cyber-resilience-act and https://digital-strategy.ec.europa.eu/en/factpages/cyber-resilience-act-implementation.