AI adoption is accelerating, but confidence is lagging
Regulatory uncertainty is now the biggest brake on AI adoption across Europe. According to AWS policy lead Sasha Rubel, 68 percent of organisations don't understand their obligations under the EU AI Act, and many are spending up to 40 percent of IT budgets on compliance.
Billy Linehan interviewed Sasha Rubel at AWS re:Invent in Las Vegas.
Europe's AI problem
AI capability remains central at re:Invent, but governance, skills and trust are increasingly part of the conversation. This message frames a policy-focused conversation with Sasha Rubel of AWS.
Regulation has become the brake
The EU AI Act is now moving from policy into practice. According to Rubel, regulatory uncertainty is now one of the biggest obstacles to AI adoption across Europe.
"One of the biggest blockers to AI adoption is regulatory uncertainty. Right now, around 68 percent of organisations don't understand their obligations under the EU AI Act, and many are spending up to 40 percent of their IT budgets on compliance-related costs."
The result is hesitation at precisely the moment when AI is moving from pilots into production. For smaller economies like Ireland, the issue is not just compliance, but speed.
Who is Sasha Rubel?
Sasha Rubel leads AI and generative AI policy for Europe, the Middle East and Africa at Amazon Web Services, working with governments and regulators on how AI is governed and deployed at scale. Before joining AWS in 2021, she spent more than a decade at UNESCO, where she led global work on artificial intelligence, digital transformation and technology policy. Rubel has personal ties to Ireland, sits on Ireland's AI Advisory Council, and has addressed the Oireachtas Joint Committee on Enterprise, Trade and Employment.
Adoption is accelerating, unevenly
AI uptake across Europe is rising, but Rubel points to a widening gap in how it is being adopted.
"What we're seeing is a two-tier economy. Startups are moving quickly on AI, while larger organisations and the public sector are moving more slowly."
The divide is not just between company sizes, but between sectors. Tech-native industries are pulling ahead, while others struggle to translate AI into practical use. For Ireland, where SMEs employ nearly 70% of the workforce, that gap isn't academic but structural.
Why the public sector matters
Rubel repeatedly returns to trust, and the role of government in building trust.
"I actually salute the Irish government, because Ireland has been very proactive in thinking through how we can accelerate AI adoption in the public sector. When the public sector leads on AI adoption, it inspires trust in citizens, and that trust really matters."
Visible, responsible public sector use of AI, according to Rubel, gives private companies confidence to move faster, particularly in highly regulated environments.
Skills, not technology, are the constraint
Many businesses say they lack in-house capability, while staff who want to upskill struggle to find the time or afford formal training. The result is slower adoption.
"There are some really key barriers to AI adoption, and the first is digital skills. Ireland has been at the forefront of thinking about how the private sector and the public sector can come together to address that challenge."
Rubel highlights EdTIPS, an AWS programme bringing AI literacy into Irish primary schools. AWS also offers free, self-paced AI training for adults, covering not just how AI works, but how to use it responsibly, including bias, fairness, privacy and security.
Rubel stresses that progress depends on collaboration.
"Ireland has been at the forefront of national upskilling initiatives that bring together government, policymakers, industry and academia, both nationally and at a local level. That kind of collaboration really matters when technology is moving this fast."
Democratising access to AI
For Rubel, skills are not just an economic issue, but a participation issue.
"We ...