Navigating the Future: The EU AI Act's Role in Shaping a Safer, Innovative, Standards-aligned World

One small step for AI Standards - one giant leap for AI Regulation

Yesterday marked a historic moment in the evolution of artificial intelligence (AI) regulation and standards, with the signing of the AI Act. This landmark legislation sets a global precedent, underscoring the European Union's commitment to leading the charge in establishing a framework that balances innovation, with trustworthiness considerations promoting safety. As we stand on the brink of this new era, it's crucial to delve into the significance of the AI Act for society, the innovation ecosystem, and the safety of AI technologies. Moreover, the role of standards in operationalizing this legislation cannot be overstated, as they are instrumental in ensuring that the Act's objectives are met effectively. Articulated beautifully by Dragos Tudorache : ".. this is not going to be like GDPR, this time we are going to mandate standards.."

The AI Act is more than just legislation; it's a visionary step towards a future where technology serves humanity responsibly. By setting clear guidelines for the development and deployment of AI, the Act fosters a safe environment for innovation. It ensures that AI technologies are developed with safety considerations at their core, promoting trust and transparency between AI entities and the public. This TRUST is essential for societal acceptance and the successful integration of AI into daily life, so it is not by chance that our WG3 in SC-42 is called Trustworthiness. Here is how trust in a legislative monumental moment looks like as described by Dr. Laura Caroli - "Even above our optimistic expectations":

Safety in AI is a multifaceted issue, encompassing data privacy, ethical AI use, and the prevention of harmful biases. The AI Act addresses these concerns head-on, establishing a legal framework that prioritizes human oversight and transparency. However, legislation alone is not enough. Operationalizing the AI Act's provisions requires the development and implementation of robust standards. These standards are the linchpins in the machinery of AI regulation, ensuring that the legislation's principles are translated into actionable guidelines that can be universally applied. Check out Barry Scannell as he touches on some key safety points through the lens of different categories such as Biometrics, Social Scoring, and more in his latest post. When it comes to safety it always needs to be balanced to support innovation - as eloquently said by Thierry Breton : "We are regulating as little as possible — but as much as needed!" (you should follow him in general as he also has great valentine day posts :))

Having a risk-based approach resonating with other global frameworks (OECD, NIST, ISO/IEC), the AI Act is set to have a ripple effect beyond the European Union's borders. Countries around the world are closely watching the EU's pioneering efforts and are likely to adopt similar measures. This global movement towards standardized AI regulation underscores the need for international collaboration in developing global standards for AI. As we navigate this uncharted territory, the role of leadership in developing these standards becomes increasingly significant. Entrusted with this leadership re standard development tailored to the EUAIA, is CEN/CENELEC's JTC21 - Led by an amazing team. A team that I had the privilege of meeting in Dublin only a few weeks ago: Chair - Sebastian Hallensleben ; Vice Chair - Patrick BEZOMBES ; Secretary - Kim Skov Hilding

In the global arena of AI standards development, leadership is not just about setting guidelines; it's about envisioning the future of technology and steering the world towards a safer, more equitable AI landscape. My journey in the realm of AI standards and trustworthiness began with a clear vision of the critical role these frameworks would play in the future of technology. Recognizing the importance of global collaboration and standardization early on, I took the initiative to register and physically attend, the inaugural meeting of ISO/IEC JTC 1/SC 42 in China in 2018, investing my resources to be at the forefront of AI standardization. This decision marked the beginning of a deeply involved and influential path in shaping the landscape of AI trustworthiness and safety via standards! This is where it all began - can you find me in the picture?

This journey and commitment continued through my contributions to the IEC SEG 10 on AI ethics where I helped lay the groundwork for trustworthy/ethical guidelines, that would inform the development and deployment of AI technologies worldwide. My work with ISO/IEC JTC 1/SC 42 further solidified my commitment to establishing a robust framework for AI standardization, focusing on creating standards, that ensure AI technologies are developed with accountability, transparency, and respect for human rights at their core; Fun Fact - back in 2018 it was not obvious at all for us to establish a trustworthiness (safety/ethics/soceital-concerns) working group...and as much as we were progressive back then coining the term... today it seems obvious to all - so I'm very proud of WG3 being established [thank you for the support Phil Wennblom ] in the first place and all the great milestones it has achieved. Here is another personal "nerdy" milestone as we gathered in Tokyo back in 2019 - do you know who/what is it next to me in the picture?

My engagement didn't stop at the international level; I also contributed to national efforts through collaborations with the SCC (Canada) and NIST (US), ensuring that the standards we developed are not only globally relevant but also adaptable to national needs and contexts. I will continue working diligently to bridge the gap between the international and national, bringing countries closer together for better collaboration; In part by contributing to the great OECD.AI core team and extended group of experts, led by Karine Perset , and especially locally in Israel as I continue my work with ???? ?????? ??????? - The Standards Institution of Israel & Israel Innovation Authority ???? ??????? , making sure Israel is at the forefront of AI, standards and regulation.

The journey from the early days of SC 42 to the ongoing work in JTC 21 reflects a continuous commitment to advancing AI standards that promote innovation while safeguarding society. This leadership journey in AI standards is not just a professional endeavor but a personal mission to ensure that we embrace the potential of AI, and we do so with a conscientious approach that benefits all of humanity. My proactive involvement from the very beginning underscores a dedication to not only participate in the conversation but also to lead it - driving the development of AI standards that pave the way for a future where technology and safety go hand in hand.

The signing of the AI Act is a monumental step forward in the journey towards responsible, safe, trustworthy AI. It represents a commitment to innovation with safety considerations in the world of AI development and deployment; While the AI Act may not be flawless, it represents a significant and commendable beginning [to learn more on this topic press here and listen to a conversation with the fantastic Kai Zenner ]. Acknowledging its imperfections is crucial, as it allows us to identify and address the gaps collaboratively. By embracing the AI Act as our guiding framework and committing to a transparent, collective effort to refine and evolve its provisions, we position ourselves to adeptly confront and leverage the myriad challenges and opportunities the future of AI holds.

As we look to the future, the role of standards in operationalizing this legislation will be crucial. Through collaborative efforts and strong leadership in the development of these standards, we can ensure that AI serves as a force for good in society. *For the non-nerds in the crowd, as you might not understand the last photo/gif in this article: standards work and regulation efforts have been siloed for too long - the EU AI ACT marks a change in bringing them closer together - a process unification by the legislator - Evolution! Now go watch the Anime movie from 1995 titled "Ghost in the Shell" & ping me if you didn't get it..or also if you did..because I can always talk about this movie.. Say Keren Elazari - when are you arranging another screening :)

As always, LMK your thoughts. Y.

Shortlist to follow on this topic:

European Commission European Parliament EU Digital & Tech

Eva Maydell (Paunova) in general and her latest post and focus on next steps.

Catelijne Muller Dan Nechita Philipp Hacker and last but not least Luca Bertuzzi