Software has become an integral part of everyday life, often operating in the background without attracting much attention. Businesses rely on applications to manage customers, process payments and store information, while individuals use connected devices and digital services for communication, entertainment and household security. Despite this dependence, users generally have limited visibility into how these products are developed, whether they undergo security testing or how long their manufacturers intend to maintain them. Such questions tend to emerge following the discovery of vulnerabilities or incidents involving compromised systems and personal information.
The European Union has been examining these concerns as part of its wider approach to digital security. One result is the Cyber Resilience Act (CRA), formally adopted as Regulation (EU) 2024/2847, which introduces mandatory cybersecurity requirements for hardware and software products made available on the European market. Rather than concentrating exclusively on how organisations respond to cyberattacks, the regulation considers security throughout the development, distribution and maintenance of digital products.
A changing regulatory environment
The Cyber Resilience Act entered into force on December 10, 2024, with its main requirements becoming applicable on December 11, 2027. The regulation covers what European legislation describes as products with digital elements, including commercially distributed software, connected devices, operating systems and certain hardware components. Some remote processing functions associated with covered products may also fall within its scope, although standalone cloud services and conventional Software-as-a-Service platforms are generally treated differently.
The European Commission identifies two recurring problems behind the legislation: insufficient cybersecurity in digital products and the lack of timely security updates. Manufacturers of covered products are consequently expected to assess cybersecurity risks, implement appropriate safeguards, maintain documentation and establish processes for identifying and addressing vulnerabilities. Security support is also addressed, with a general minimum period of five years unless a shorter expected product lifetime justifies an exception.
These requirements may have different implications depending on the manufacturer. Larger technology companies may already have dedicated security teams and established vulnerability management procedures, while smaller developers may need to introduce processes that previously received limited attention. The distinction is particularly relevant where software is developed commercially but maintained by only a small number of people.
Reporting vulnerabilities becomes mandatory
An important implementation milestone arrived on September 11, 2026, when the CRA’s reporting obligations began applying to manufacturers of covered products. Under these rules, manufacturers must report actively exploited vulnerabilities and severe incidents affecting product security. An early warning must be submitted within 24 hours of becoming aware of a reportable issue, followed by a more detailed notification within 72 hours.
These reports are submitted through a central platform established by the European Union Agency for Cybersecurity, known as ENISA, allowing relevant national cybersecurity authorities to coordinate their responses. The requirements do not apply to every programming error or minor vulnerability, but to specific categories of incidents and actively exploited weaknesses defined by the regulation.
The emergence of AI-assisted software development
The introduction of the CRA also coincides with considerable changes in software development. Artificial intelligence, automated coding assistants and increasingly accessible development platforms have made it possible to create applications with fewer resources and, in some cases, less traditional programming experience. These developments are attracting interest from established software companies as well as individuals and smaller businesses exploring commercial applications.
One increasingly familiar approach is vibe coding, where developers use natural-language instructions to generate substantial portions of an application. While these tools can accelerate development and make experimentation more accessible, they also raise questions about how generated code is reviewed and whether those deploying it fully understand its underlying architecture.
An application can appear to function correctly while containing weaknesses that are difficult to identify through ordinary testing. A customer management platform, for example, might provide a polished interface and working authentication while exposing information through improperly configured database permissions or insufficient access controls. These issues are not exclusive to AI-generated software, but automated development can make the distinction between functional software and securely implemented software particularly relevant.
The CRA does not establish separate obligations for AI-generated code. Its requirements depend on the characteristics of the product and the responsibilities of the manufacturer rather than the tools used during development. For businesses building commercial software with AI assistance, this distinction may become increasingly important as they consider how testing, maintenance and security responsibilities should be organised.
The implications for smaller businesses
For independent developers and small technology companies, cybersecurity requirements may also influence the economics of software development. Creating an application involves an initial investment, whereas maintaining its security can require ongoing resources long after the product reaches customers. Vulnerabilities discovered in third-party libraries, development frameworks or operating environments may necessitate updates even when the product itself has otherwise remained unchanged.
The European Commission has acknowledged the difficulties smaller organisations may face. In July 2026, it published additional implementation guidance containing 67 practical examples, alongside explanations of product scope, support periods, substantial modifications and reporting requirements. The guidance specifically considers microenterprises and small and medium-sized businesses, although the interpretation of individual obligations will still depend on the products concerned.
The Commission described the guidance as an attempt to make implementation more accessible, particularly for companies without extensive regulatory resources. Whether smaller developers will find compliance proportionate in practice may become clearer as technical standards develop and national authorities begin applying the requirements more extensively.
Beyond regulatory compliance
A recurring consideration in cybersecurity regulation is the distinction between documenting security practices and achieving meaningful security outcomes. Technical documentation, formal assessments and established procedures can help demonstrate that risks have been considered, although they cannot guarantee that a product is free from vulnerabilities. Even extensively tested software may require further updates when new weaknesses are discovered or the surrounding technical environment changes.
The CRA attempts to address this uncertainty by considering cybersecurity throughout a product’s lifecycle rather than exclusively at the point of sale. Its long-term effectiveness will depend partly on how consistently manufacturers implement these requirements and how supervisory authorities interpret them across different categories of products. For businesses purchasing software, the regulation may eventually provide clearer information about security support and manufacturer responsibilities. For developers, it introduces additional considerations into decisions about product architecture, maintenance and commercial viability.
As December 2027 approaches, the practical implications of the legislation are likely to receive greater attention across Europe’s technology sector. At the same time, the growing accessibility of software development raises a broader question about how technical expertise, commercial responsibility and cybersecurity expectations should evolve together. The Cyber Resilience Act offers one regulatory approach to that question. How it will influence the software market may become more apparent over the coming years.
Sources and further reading
European Union — Regulation (EU) 2024/2847
The official legal text establishing the Cyber Resilience Act.
European Commission — Cyber Resilience Act
Overview of the regulation, scope and implementation timeline.
European Commission — CRA Reporting Obligations
Reporting deadlines and requirements effective September 2026.
European Commission — CRA Implementation Guidance
Guidance published July 27, 2026, including practical examples for smaller businesses.
ENISA — The CRA Single Reporting Platform Is Launched
September 11, 2026 press release.





