UNECE launches declaration on products with embedded AI calling for global cooperation to address regulatory challenges

In response to the rapidly growing complexities of artificial intelligence (AI) and digital technologies embedded in everyday products and services, the UNECE Working Party on Regulatory Cooperation and Standardization Policies (WP.6) has issued a declaration aimed at fostering increased global regulatory coherence. This landmark initiative addresses the urgent need for a coherent, international approach to regulating AI embedded in products, such as consumer electronics, medical devices, and industrial systems and machinery.

The declaration builds upon the related Overarching Common Regulatory Arrangements (CRA), a framework designed to support the development of voluntary regulatory cooperation between governments, while safeguarding global trade and technological advancements. The CRA provides guidance on reducing technical barriers to trade, aligning regulatory approaches, and enhancing market surveillance.

“The fragmented regulatory landscape for AI, combined with global security concerns and geopolitical challenges, risks creating new barriers to trade. It is thus vital that digital risks are addressed with measures that are effective for the specific AI use cases,” said Heidi Lund, the Chair of UNECE WP.6. “This declaration underscores the importance of consensus-driven regulatory approaches to counter digital vulnerabilities and ensure that technological advancements benefit all economies.”

The key benefits of the declaration include:

• Facilitating International Trade: By promoting increased regulatory coherence, the declaration helps to eliminate redundant checks and supports smoother trade across borders.
• Enhancing Regulatory Coherence: Government agencies adopting the CRA gain greater clarity and predictability, avoiding unnecessary technical barriers while promoting mutual recognition.
• Promoting Trustworthiness and Resilience: The CRA addresses the societal and technological risks associated with AI, ensuring that embedded AI systems are safe, secure, and resilient to cyber threats.
• Supporting Sustainable Development: The CRA aligns with the UN Sustainable Development Goals (SDGs), encouraging AI implementations that advance environmental, economic, and social objectives.

The declaration represents a proactive response to the rapidly evolving digital landscape. It provides a unique platform for global regulatory collaboration, offering a flexible, non-binding methodology that adapts to the maturity levels of participating agencies. This flexibility ensures that agencies can implement best practices tailored to their specific regulatory needs, without the burden of binding commitments. The end goal is to ensure that products entered onto the market are safe for consumers and that the AI elements do not negatively impact human autonomy, mental wellbeing or individual freedoms.

The UNECE W.P.6 Project Leader, Markus Krebsz, emphasized the broader impact of the CRA, noting that it “not only facilitates trade and regulatory coherence but also addresses key societal challenges like trustworthiness, privacy, cybersecurity and reducing the global digital divide. The framework encourages continuous compliance and the sharing of best practices to ensure the safe and responsible development of AI technologies.”

Next steps for governments and agencies: Call for engagement

As the declaration moves forward, UNECE will soon launch a call for engagement, inviting governments and national agencies to participate in new implementation projects, aimed at bringing the principles of the CRA to specific sectors, such as medical devices, electrical appliances and/or toys. Governments and national agencies globally are invited to join this initiative, which offers a strategic path to shaping the future of AI regulation while contributing to a more integrated and efficient global market for AI-embedded products and services.

For more information, please visit this page or see the UNECE WP.6 website, or contact [email protected].