Regulation must be an enabler, not an obstacle to unlocking the potential of AI, the William Fry AI Summit 2026 (14 May) has heard.
In his keynote speech to the 400 attendees, Michael McGrath (EU Commissioner for Democracy, Justice, the Rule of Law and Consumer Protection) addressed the question of how geopolitics was shaping AI implementation in Europe.
“It is how we govern AI within the EU that will shape Europe's geopolitical position, because our digital sovereignty ultimately rests on our democratic sovereignty,” he said.
Shape or be shaped
Commissioner McGrath observed that AI systems had the capacity to shape democracy: "If democratic societies do not shape AI through law, institutions and public accountability, then the converse is true: AI will shape us."
This will be ensured by measures such as the European Democracy Shield, the European Centre for Democratic Resilience, the Digital Services Act and the AI Act.
The AI Act, which the commissioner described as “a world-first governance framework”, will deliver “clarity and certainty for businesses across the single market, strengthening Europe's global economic clout”.
The commissioner outlined the EU's commitment to projecting its rights-based approach beyond its borders through a network of ‘adequacy arrangements’.
“We are ensuring that when data moves, rights travel with it and, with our international partners, we are showing how the protection of personal data is not a barrier to trade, but actually the foundation of a sustainable digital economy.
Global gold standard
“These agreements are a global gold standard, signalling those who have them as safe and mature destinations for digital investment.”
The EU Commissioner added that “our democratic values are not a constraint in a competitive world. They are, in fact, a strategic advantage. We do not have to choose between innovation and values.
"Europe's strength is that we insist on both."
The EU must position itself as "a developer and deployer, a leader of AI and AI infrastructure," with 15 AI factories expected to become operational under the EU's AI Continent Action Plan before the end of 2026.
A panel discussion, chaired by Dr Barry Scannell (William Fry) followed.
Sasha Rubel (Amazon Web Services) reported that while more than 54% of European businesses had now adopted AI, with 4.4 million new adopters in the past year, over 60% remained at the most basic stage.
“There's a two-tier economy that's emerging, where start-ups are steaming ahead in AI adoption and larger enterprises are lagging behind,” Rubel explained.
This matters, she explained, because “moving to advanced AI adoption could unlock €191 billion in GPA, which is part of a bigger pie of €600 billion between now and 2030, so it's a key economic lever”.
It was really concerning, she added, that “four in ten start-ups would consider leaving Europe because of better access to funding, clear regulation, and market access. That goes to over half for the most advanced start-ups that are deploying agentic AI.”
Rubel also reported that 68% of businesses did not understand their obligations under the AI Act, and that 42% of EU IT spend was consumed by compliance costs, against 22% in Japan.
The IMF had estimated this to be the equivalent of a 110% internal tariff on innovation, she said, noting that 41% of businesses identified internal EU fragmentation as a direct barrier to scale.
Loss of start-ups
Commissioner McGrath said the EU Inc proposal, “the most fundamental overhaul of EU company law in its history”, aimed to tackle the loss of start-ups.
He said: “We very deliberately chose a regulation as the legal instrument rather than a directive” to avoid the necessity of interpretation and transposition by 27 different states.
It will be “an optional additional regime” to national regimes and the “fully harmonised set of company-law features across the European Union” will include:
Talent
Crucially, this will be very attractive for capital…[and] “will make attracting and retaining talent in Europe, I believe, much easier,” he said.
The Digital Omnibus, he added, aimed to simplify regulation.
Malcolm Byrne TD (Joint Oireachtas Committee on AI chair) said that Ireland’s new AI Office could not be a ‘thou shalt not’ regulator.
“It's intended to be is an enabling role, to be able to provide support to existing regulators” and to businesses.
Byrne also highlighted Ireland’s role as a bridge between transatlantic partners.
Speaking of the role of the committee Byrne said: “My view is the technology is not a goal in itself. It's an enabler.
“So our core question is: if we're looking to better deliver public services, if we're looking for government to be able to provide services more effectively and efficiently to our citizens – can we use particular tools to do that?”
Sandboxing
He cited the work already done with AI in the Courts Service, “sandboxing” AI use for translation and stenography, and the Naval Service using AI to monitor Ireland’s maritime economic area.
Rubel said that AI literacy would be needed in more than 80% of jobs in the five years and, despite lack of AI skills in the workforce being “the biggest blocker” to AI implementation, skills building had not increased over the past year.
Deputy Byrne said “skills in the age of AI” would be a focus of the Government-hosted AI summit on 14 October.
Disruption in employment
“One of the geopolitical trends we are going to see is huge disruption in employment”, he said.
“We're going to see a lot of tech jobs go quite soon, but there'll be a lot more created.”
“It's not MAGA, it’s MECA – Make Europe Competitive Again,” Byrne added.
While there was a need to educate, upskill, and re-skill, there were some “great possibilities for Ireland and Europe”, he said.