FII Priority Europe 2026: Saudi-Led Summit Declares AI Sovereignty Requires Owning the Full Stack

Sovereign AI Means Owning the Stack, Not Just Writing the Rules

The Future Investment Initiative's Rome summit concluded on 19 June with a message aimed as much at Gulf capitals as at European ones: true artificial intelligence sovereignty requires direct ownership of chips, compute, energy, and data infrastructure, not regulatory frameworks alone. The two-day FII Priority Europe 2026 gathering, held under the theme "Europe Reimagined: Capital, Sovereignty and Strategic Autonomy," brought together global investors, policymakers, and technology leaders to examine how nations can build durable positions in the AI economy rather than simply regulating the technology others build.

The summit was opened on its final day by H.R.H Princess Dr. Maha bint Mishari bin Abdulaziz Al Saud, who assumed the role of CEO of the FII Institute earlier this year, making her first plenary address to the summit. She called on governments, investors, and business leaders to treat the frameworks developed in Rome as binding commitments rather than aspirational roadmaps.

The Sovereignty Question

The sharpest framing of the summit's central theme came from Aidan Gomez, CEO and Co-Founder of Cohere, who defined the threshold for genuine AI sovereignty in blunt terms:

"Countries need capabilities that no one else can mess with, see the data inside of, or switch off at will. That is what true AI sovereignty means."

Gomez's argument, that sovereignty requires controlling the full technology stack rather than licensing access to it, resonated with a concern running through the entire GCC region's AI strategy conversations. The UAE's Stargate initiative, Saudi Arabia's National Data Centre programme, and MGX's pursuit of direct data centre ownership are all expressions of the same logic: influence over AI infrastructure cannot be achieved through minority investment or regulatory preference alone.

The Inclusive Dimension

The summit also surfaced a harder question about who sovereign AI actually benefits. Deemah AlYahya, Founding Secretary-General of the Digital Cooperation Organization, warned that the global AI economy risks deepening existing divides if sovereignty becomes a concept reserved for wealthy nations with the capital to build full AI stacks independently.

AlYahya noted that the AI market is projected to grow from $189 billion in 2023 to $4.8 trillion by 2033, yet fewer than one third of developing countries currently have national AI strategies, and around 118 nations remain underrepresented in global AI governance structures. Her prescription was not for every country to build independent AI ecosystems, but for new models of digital cooperation based on shared infrastructure, talent mobility, and cross-border knowledge exchange.

The DCO, headquartered in Riyadh, has positioned itself as the coordinating body for exactly this kind of cooperation across its member states, many of which span the MENA and Gulf region.

Regulation as Bottleneck

Travis Kalanick, CEO of Atoms and founder of Uber, added a pointed warning about the cost of regulatory delay that applies as directly to GCC markets as to Europe:

"When the regulatory process becomes the outcome, something that is genuinely good for people can roll out a decade late. That is how innovation gets delayed even when the benefits are already clear."

The observation matters for the region because several GCC governments are currently developing AI regulatory frameworks that will govern how models are deployed, how data is stored, and how liability is assigned. Getting that architecture right without creating the kind of regulatory drag Kalanick described is one of the defining policy challenges of 2026 for markets like the UAE and Saudi Arabia.

What Rome Means for the Gulf

The FII Institute is a Saudi-founded body, and its summits are rarely purely European conversations. The sovereign AI debate in Rome is a direct reflection of the strategic questions Gulf governments are already working through at home: whether to own compute or lease it, whether to build national models or deploy foreign ones under licensing agreements, and whether regional cooperation on AI infrastructure is more effective than independent national programmes.

The UAE's position in this race has already moved well beyond ambition. From the 4-trillion transistor chip unveiled at the World Government Summit to the 5-gigawatt Stargate UAE campus under construction in Abu Dhabi, the infrastructure decisions being made today are concrete, not aspirational.

The conclusions from Rome offer a clear directional answer for the broader region. Own the stack, compress procurement timelines, anchor governance in independent institutions, and ensure that regulation enables deployment rather than delaying it. For a region that has committed hundreds of billions of dollars to AI infrastructure over the next decade, that is less a roadmap and more a confirmation of where the strategic logic already points.