The AI Vibe-Shift Comes to Europe

I spent the past week in Paris at the AI Action Summit, and came away with three observations about where the global AI conversation is heading. None of them are particularly reassuring.

1. The New Mantra: AI Adoption

The dominant theme in Paris wasn't whether AI can be built safely, or even whether it should be. That conversation has largely been settled — at least in the minds of the industry. The new mantra is adoption. How quickly can governments, businesses, and institutions integrate AI into their operations? How do we remove friction from deployment?

This reframing carries real consequences. When the conversation shifts from "should we" to "how fast can we," it creates a momentum that risks steamrolling the normative debates we still need to have. Questions about rights, accountability, labour displacement, and democratic oversight get recast as implementation details rather than foundational policy choices. When inevitability replaces deliberation, crucial decisions get made by default rather than by design.

2. From Governance to Foreign Direct Investment

At the earlier AI summits — Bletchley Park and Seoul — the central themes were risk and safety. Governments were grappling with how to build regulatory frameworks for a fast-moving technology. In Paris, that conversation was largely replaced by a different one: how to attract investment. Country after country positioned itself as a destination for AI capital. The question wasn't how to govern AI, but how to lure the companies building it.

This shift has been accelerated by the new US administration's open hostility to digital regulation. With Washington signalling that it views AI governance as an obstacle to competitiveness, other governments have taken the cue. Risk assessments and safety research are increasingly being relegated to civil society organizations and academic groups — important actors, but ones without the regulatory authority to actually enforce anything.

The result is a governance vacuum at precisely the moment we need institutional capacity the most.

3. The First Example of "Smash and Grab" US Diplomacy

Vice President Vance's appearance at the summit offered a preview of what American AI diplomacy looks like under this administration. The approach: arrive, derail the conversation, release a disruptive statement, and leave. Vance used the platform to signal that the US would not support international AI safety regulations and that it views such efforts as impediments to American innovation and dominance.

This matters beyond the summit itself. It signals a broader threat to digital governance regimes around the world. Canada's Digital Services Tax, the EU's GDPR, Digital Services Act, and Digital Markets Act, the UK's Online Safety Act — all of these now exist under a cloud of US pressure. The message from Washington is clear: regulation is a barrier, and allies who pursue it can expect pushback.

What Comes Next

Kate Crawford put it well when she described AI as being in its "empire era." What I saw in Paris confirmed that framing. The AI realignment currently underway is not primarily about technology. It is about power — specifically, about consolidating US technological and economic dominance at the expense of multilateral governance.

The question for countries like Canada, and for institutions trying to build durable AI policy, is whether they will hold the line on the governance frameworks they've been developing, or whether they'll be swept along by the adoption imperative and the geopolitical pressure that comes with it. The answer to that question will shape the digital order for decades to come.