Artificial intelligence, software and connected digital functions are fundamentally changing the requirements for product safety and product liability. For modern products, safety no longer depends solely on the quality of physical components. Software architecture, data quality, automated decision-making, cybersecurity, remote maintenance and security updates can be equally decisive in determining whether a product is safe and meets the safety expectations that users are entitled to have.

This is particularly relevant for automated and autonomous vehicles, but also for machinery, medical devices, robotics, industrial installations and connected consumer products. In these sectors, hardware, software, sensors, cloud and back-end services, and AI-enabled functions interact closely. Defects may therefore result not only from a faulty physical component, but also from defective algorithms, insufficient testing, cybersecurity vulnerabilities, problematic data, or a missing or inadequate software update.

New EU Product Liability Directive

The new EU Product Liability Directive is an important step in adapting product liability law to the digital economy. EU Member States must transpose it into national law by 9 December 2026. The Directive clarifies that software, including AI systems, is covered as a product under the strict, no-fault product liability regime. It reflects the fact that digital functions and connected services are now central to the safety of many products.eur-lex.europa+1

For companies, this increases the importance of a comprehensive product liability and product safety strategy. Product safety must be addressed throughout the entire product life cycle: from concept and development, testing and validation, through market entry and product monitoring, to cybersecurity, security updates, warnings and recall measures.

Safety by Design and Secure Updates

The evolving legal framework creates clear incentives for Safety by Design and Security by Design. Companies should address safety and compliance requirements at the development stage of hardware, software and AI-enabled functions. This includes documented risk assessments, clearly defined development and approval processes, traceable decision-making, careful testing, effective vulnerability management and clear allocation of responsibilities among manufacturers, suppliers, software developers and platform providers.

Secure software and over-the-air (OTA) updates are particularly important for automated and autonomous vehicles. A carefully developed, tested and documented security update can promptly close a vulnerability across a vehicle fleet and effectively reduce risks in the field. Depending on the nature and level of the risk, such measures can be a useful complement to – or an alternative to – conventional recalls requiring vehicles to be brought into workshops.

Chambers Contributor – Practice Guide: Further information is available in the Chambers Global Practice Guide: Product Liability & Safety 2026 – Austria.practiceguides.chambers

Product Liability Advice

EUSTACCHIO advises manufacturers, suppliers, software companies, platform providers and operators of connected systems on product liability, product safety, product compliance and cybersecurity. Our advice includes the legal assessment of AI and software functions, the design of safety and compliance processes, contractual arrangements throughout the supply chain, the preparation of security updates and recalls, and support in connection with product safety incidents.