Major Overhaul to EU Product Safety Regulations

A significant update to the European Union’s product safety framework has come into effect, requiring companies to assess the safety of products using artificial intelligence (AI) for their entire lifespan. The new rules, introduced under the EU General Product Safety Regulation (GPSR), emphasize the evolving nature of AI technology and mandate safety assessments that cover products from development to updates and beyond.

Key Aspects of the New Regulations

As of December 13, the GPSR requires that any product incorporating AI undergo safety evaluations that address its ongoing lifecycle, including changes made through updates or modifications over time.

Safety Considerations Under the New Framework

Under the new rules, companies must now evaluate several factors in their safety assessments, including:

• Mental health impacts.
• Cybersecurity vulnerabilities.
• The effect of software updates and any software that may be downloaded to the product.
• AI and machine learning functionalities.
• The specific health and safety risks posed by products to different genders and vulnerable consumers.

Implications for Companies Using AI

Experts also caution that companies customizing AI for their own applications could be classified as "manufacturers" under the new GPSR. This designation would hold them accountable for safety assessments in the same way as the tech companies that originally developed the AI. This could significantly expand the scope of responsibility for businesses that integrate AI into their products.

Conclusion: Broader Responsibilities for AI Product Safety

With the GPSR now in effect, businesses using AI must carefully assess the long-term impacts of their products, factoring in everything from software updates to potential risks to vulnerable groups. These comprehensive safety assessments are critical for ensuring the responsible use of AI technology within the EU and protecting consumers from evolving risks.