The Trilogue Breakdown: What Just Happened

Negotiations between the EU Council, European Parliament, and Commission over AI Act reforms collapsed this week, with Cyprus—currently holding the rotating EU presidency—unable to broker agreement on how to classify high-risk AI systems embedded in medical devices and toys.

The sticking point wasn’t philosophical. It was technical: regulators disagreed on whether sectoral legislation (medical device rules, toy safety standards) should supersede or sit alongside the AI Act’s Annex III high-risk classifications. This might sound procedural, but it’s existential for Ireland’s August 2026 deadline.

Why Ireland Is Uniquely Exposed

Unlike most EU member states, Ireland didn’t establish a standalone national AI regulator. Instead, the Irish government empowered 15 competent authorities across different sectors—health, finance, transport, telecoms—to enforce AI Act compliance within their domains.

This distributed model made sense on paper: sectoral experts handling sectoral risks. But it creates a coordination nightmare when the sectoral legislation itself is in flux. If Annex III definitions shift, each of Ireland’s 15 authorities needs to understand and implement those shifts consistently. Right now, with trilogue negotiations frozen, that guidance doesn’t exist.

A Dutch MEP called the situation “regulatory chaos for European companies that prioritised safety compliance.” For Irish companies subject to 15 different enforcement regimes, it’s worse than chaos—it’s strategic paralysis.

The Practical Problem for Irish Builders

High-risk AI systems embedded in medical devices and industrial equipment must comply by August 2, 2026. That’s approximately 100 days away. Companies building these systems need to know:

  • What documentation is actually required?
  • Who inspects it—the medical device authority or the AI authority?
  • If requirements conflict, which takes precedence?

Right now, the answer from Dublin is: we don’t know yet. And the answer from Brussels is: we’re negotiating that.

Companies that invested in AI Act compliance based on January 2026 guidance may now discover their approach was built on assumptions about sectoral precedence that the Parliament rejects. Rework, delay, and regulatory exposure loom.

What Happens If Talks Don’t Resume Successfully

If trilogue negotiations fail entirely, the original August 2026 deadline for Annex III compliance stands unchanged. But the standards, technical documentation frameworks, and guidance that make compliance possible don’t exist.

Ireland’s AI Office—mandated to be operational by August 1—would launch into a legal and technical vacuum. The 15 competent authorities would have no coordinated framework for risk assessment, testing protocols, or compliance audits.

Open Questions

When do negotiations resume? The May trilogue round is the critical juncture. If Cyprus can’t broker agreement, the presidency passes to Hungary in July—leaving one month to fix months of disagreement.

Does the Commission issue interim guidance while negotiations stall? Will Ireland’s AI Office provide sectoral guidance unilaterally? Or does the market simply freeze, with high-risk deployments delayed pending regulatory clarity?

The stakes for Irish medtech firms, AI-driven financial platforms, and industrial automation vendors are enormous. So far, the conversation in Dublin has focused on administrative readiness. The real crisis is regulatory readiness—and that’s out of Ireland’s hands.

What’s your high-risk compliance timeline looking like? Track the trilogue resumption schedule closely.

Source: EU Council (Cyprus Presidency)