Yes, #Europe is moving in the right direction. But lack of alignment between ambition and execution risks leaving competitiveness stuck in between.

#DigitalEurope #EUcompetitiveness

On sovereignty, Viola was rather blunt, calling the EU’s reliance on foreign #cloud providers “ridiculous.” He urged greater trust in European providers and a coordinated approach to #RadioSpectrum, to be treated as a common EU resource for investment and #security.

On industrial scale-ups, the promise of easier #mergers remains valid in theory, but only the forthcoming merger guidelines will show whether practice follows.

His defence? Reasoning #AI is barely “one year old,” #GPUs remain costly, and many smaller firms cannot yet afford the leap. He placed his hopes on the upcoming #ApplyAIStrategy, with #healthcare as its first testing ground.

Europe’s race for #DigitalCompetitiveness was at the center of the discussions, and Roberto Viola (DG CONNECT) acknowledged one crucial issue. #AI adoption in Europe is still painfully low.

Today, the #FT Connect Europe Forum represented a great opportunity to not only take stock of the state of play, but also capture the mood around #EU policy-making. Put simply, the gap between policy-makers and industry seems to persist.

#Europe must not be forced into a false choice between protecting its citizens and competing globally.

It must find a way to bridge the two.

But the other part of this story is that #Europe cannot regulate alone. #AI suffers from a tragedy of the commons. The solution is to regain the position to set standards and level the playing field. The #BrusselsEffect, anyone?

It is profoundly sad that #Europe, the continent most committed to making #AI safe and socially responsible, may now have to step back, not because it wants to, but because it has no other option if it wishes to remain a global power able to set standards later.

And in a world where technological advancement, especially through #AI, is tied to both military and economic strength, this weakness matters. Negotiating power on the global stage is built on precisely those two elements.

It is about Europe’s strategic position. In 2024 the #US produced 40 notable #AI models, #China 15, and #Europe only 3. There's no way around it anymore. We are lagging.

Where Draghi’s words become most revealing is on high-risk systems. He calls for pausing the implementation of the #AIAct in this area until we “better understand the drawbacks.” But this is not only about understanding drawbacks.

Voluntary codes are no guarantee of compliance, enforcement remains under-resourced, and new forms of manipulation or disinformation will test the #AIAct limits for unacceptable risk-systems.

One year after the #Draghi report, his remarks on the #AIAct reveal Europe’s dilemma. Yes, the ban on unacceptable-risk systems landed without major resistance, thanks to codes of practice and the Commission’s guidelines. But that's not where this story ends.

But SMEs don’t necessarily need to follow a step-by-step path through predictive and generative AI. As Professor Angela Aristidou suggested, the integration of earlier AI systems into workflows has been difficult. Instead, SMEs may be able to leapfrog directly into the #AgenticAI era.

As Michelle Gill from PayPal explained, the biggest opportunity of #AgenticAI is that it can level the playing field so SMEs can compete with big corporates. But this requires SME-friendly tools, fair access to AI-driven online discovery, and proportionate regulatory frameworks.

Some of the biggest pain points for SMEs. Time is scarce and #AgenticAI can take over entire workflows, freeing up hours to be used for scaling. Trust must be embedded from design through deployment. Talent is also about mindset and soft skills, such as knowing when human intervention is required.

Yet the divide is already clear: adoption by large companies sits at 40%, while SMEs lag at just 10%. The risk of a widening digital divide is real. But so too is the opportunity to make AI a democratizing force, if we get the frameworks for trust, risk and interoperability right.

Across OECD countries, SME adoption of #AI has nearly doubled since 2023. But when it comes to agentic AI, we’re still at the starting line. That’s because agentic AI has autonomy, long-term memory, and an expansive feedback loop. That marks a true paradigm shift from reactive to proactive systems.

“A ship in the harbour is safe, but that’s not what ships are built for.”

That’s how today’s OECD #Digital4SME webinar on #AgenticAI and what it means for #SMEs opened. A fitting reminder that while caution feels safe, growth demands venturing into new waters.