Standards rarely make headlines. Yet they are among the most powerful tools for translating policy ambitions into practical solutions. Nowhere is this more evident than in the European AI ecosystem, where standards are expected to play a critical role in supporting the implementation of the AI Act.
As Secretary of CEN-CENELEC JTC 21 Artificial Intelligence, I've had the privilege of watching this process unfold from the inside. It has been challenging, demanding, and deeply rewarding and it has given me a front-row seat to how Europe builds consensus around emerging technology.
Most people know standards like ISO 9001 or the various cybersecurity standards. Far fewer understand how standards come into being.
The European standards development process rests on a simple but powerful principle: stakeholder consensus.
Unlike legislation, standards aren't written by governments alone. They're developed by representatives from industry, academia, public authorities, consumer organisations, NGOs, and many other groups. These experts come together to discuss, debate, and ultimately agree on what good practice looks like in each domain.
The process typically starts with a recognised need — arising from technological developments, market demand, regulatory requirements, or societal challenges. Once a work item is approved, experts are appointed to working groups and drafting begins.
From there, the draft is refined through rounds of technical discussion, internal review, public consultation, and formal voting. National standards bodies coordinate stakeholder input in their own countries and bring consolidated positions into European discussions.
Reaching consensus can take time, especially with complex, fast-moving technologies. But that's also one of the greatest strengths of standardisation. Standards shaped this way tend to reflect diverse perspectives, and end up more practical, balanced, and broadly accepted than anything a single organisation could produce on its own.
For a technology like AI, that matters enormously — it's what allows technical expertise, societal concerns, and regulatory expectations to be weighed together, rather than in isolation.
Against this backdrop, CEN and CENELEC established Joint Technical Committee 21 (JTC 21) on Artificial Intelligence in March 2021. The committee was created to give Europe a dedicated platform for developing AI standards, and it later became the focal point for delivering the standards supporting the European Commission's standardisation request linked to the AI Act.
Today, experts from more than twenty European countries take part in JTC 21's work. They represent large enterprises, SMEs, universities, research institutions, government authorities, consumer representatives, and civil society organisations.
The committee's mandate spans a wide range of AI-related topics like risk management, quality management, trustworthiness, cybersecurity, datasets, bias, conformity assessment, and more, all relevant to implementing the AI Act.
Most importantly, JTC 21 bridges regulatory requirements and practical implementation. Legislation alone is often too high-level to guide technical work. Standards fill that gap, offering the detailed specifications, methodologies, and guidance organisations need to actually follow through.
When the opportunity arose for Danish Standards to take on the JTC 21 Secretariat, I saw it as a rare chance to contribute directly to one of Europe's most significant digital policy initiatives.
The AI Act is an ambitious attempt to regulate a transformative technology while still protecting innovation and competitiveness. But legislation alone can't achieve that balance, standards are a crucial part of how it gets done.
That intersection of technology, policy, governance, and stakeholder engagement was one of the things that drew me to this role. It's rare to find a position that combines all of these dimensions while having a direct hand in shaping how organisations across Europe will meet future requirements.
For Danish Standards, taking on the Secretariat also carries strategic weight within the European standardisation landscape. It lets us facilitate discussions, support collaboration among stakeholders, and help ensure European expertise is properly reflected in the standards being developed.
When people hear "secretary," they often picture administrative support. In reality, the role is much broader than that.
The Secretariat serves as the committee's operational backbone by coordinating meetings, supporting working groups, facilitating communication, managing processes, ensuring procedural compliance, and maintaining momentum across a genuinely complex portfolio of projects.
A large part of the job is bringing people together.
JTC 21 brings in experts with very different backgrounds, priorities, and experiences. Creating a space where these perspectives can be discussed constructively is essential to reaching consensus.
The Secretariat also acts as the connecting point between committee leadership, working groups, CEN-CENELEC management, the European Commission, national standards bodies, and a long list of external stakeholders. In this role, coordination matters as much as technical knowledge.
One of the most rewarding parts of this job has been seeing the sheer level of engagement across Europe.
AI moves fast, and views on it vary widely. Bringing together experts from different countries and sectors creates an environment where knowledge-sharing happens constantly. Every meeting brings new insight into emerging technology, implementation challenges, and regulatory thinking.
This role has also given me opportunities to strengthen collaboration between national standards bodies and to build wider awareness of the value standards provide.
Over the past few years, there's been a growing recognition that standards aren't just technical documents. They're strategic tools — for innovation, market access, trust, interoperability, and regulatory compliance alike.
Watching that understanding take hold across the AI ecosystem has been genuinely encouraging.
Of course, this journey hasn't been without its difficulties.
The first challenge is speed. AI technology moves incredibly fast, while standardisation, by design, demands careful consideration, consultation, and consensus-building. Balancing quality and inclusiveness with the need for timely delivery creates constant tension.
The second challenge is complexity. Many of the topics JTC 21 addresses involve technical, ethical, legal, and societal considerations all at once. Reaching consensus among such a diverse group of stakeholders takes patience, diplomacy, and a genuine willingness to understand other perspectives.
The third challenge is scale. The AI Act has generated unprecedented interest in AI standards. Expectations from policymakers, regulators, industry, and the public are high — and managing those expectations while running a large, ambitious work programme takes serious coordination and collaboration.
As Europe continues implementing the AI Act, standards will only grow in importance.
The standards coming out of JTC 21 are meant to help organisations understand requirements, build effective governance structures, reduce risk, and develop trustworthy AI systems.
For me personally, this role has reinforced a simple lesson: meaningful progress rarely happens through individual effort alone.
What makes standardisation unique is its ability to bring competitors, regulators, researchers, and civil society to the same table to solve shared problems. In a field as transformative as artificial intelligence, that collaborative approach matters more than ever.
The work is far from finished, but it's exciting to help shape a framework that may influence how Europe develops and deploys trustworthy AI for years to come.
