The Artificial Intelligence Act, adopted by the European Union last month, has raised some interesting questions for me about the symbiotic relationship between innovation and regulation.

There are obviously a host of very sensible measures within the Act; the efforts to ensure safety, security, and ethical behaviour should be welcomed at a time when the concept of “trust” in AI is still being debated.

My area of concern centres more on the disparity that I see in the EU’s dominant focus on regulation and governance rather than the innovation of the technology itself.

The UK and EU appear to be in the midst of a “regulatory one-upmanship” battle, with the Act’s approval another notable milestone in that. Yet this perhaps distracts us from the huge promise within these territories – France has emerged as a culturally inventive hub for AI, while the UK boasts world-class tech-focused universities packed with intellectual power.

The pendulum may swing the other way in time, but at this moment, I observe that the EU would rather push to be a global leader in regulation when I wish it were creating more to be regulated.

Success isn’t a purely financial metric

Absent a reversal in this polarity, the US will continue to lead the way. It will use its capital structures, leverage large tech universities, and draw upon its huge tech sector to further assert global dominance in AI. It will likely opt to use its existing legal structure to regulate, perhaps with some provisions updated over time.

If we were to measure success in purely financial terms, the US will, therefore, “win”. However, we must remember that virtual entities like AI don’t live within geographical boundaries. Last summer, I read some of the underlying research papers on GenAI and was struck by the level of international collaboration they revealed, often transcending geopolitical tensions.

Therefore, comparing regional entities in a notional race misses the fact that virtual technologies can be used across borders no matter where the technology is built, and innovation reflects a global science community that is not yet defined by politics.

As such, perhaps we should be asking a different question: which countries, regions, organisations, and individuals are offering the most towards the development of safe and effective AI, and who is deriving the most from these opportunities? That is the race that really matters.

AI is already at work – off the grid

We must not forget that forging ahead in AI is not the exclusive privilege of only Western nations and China. There are versions of GenAI being deployed using small language models on non-smartphones in many developing countries or off the grid entirely without internet connectivity. 

This is possible because once a language model is trained using the huge computational power required, it can become a modestly-sized programme for localised use – the same as everyday office computer applications.

The best session I attended at the World Economic Forum in Davos this year was one that convened a number of social entrepreneurs. They were using the technology in this way for real; they were well into implementation, extending the footprint of AI to enhance healthcare access in Brazil, for example, bringing further credence to the point that its virtual nature defies geographic boundaries and, to some extent, jurisdictional policies. 

If AI follows other technology innovation waves, private capital will determine the evolving AI ecosystem much more than public entities. But, for now, let’s take some comfort in the fact that AI cannot be boxed in by borders, politics, or even regulation. It may even be the first technology wave to solve more problems of access to digital benefits than it creates.