Contact us Dublin · Lisbon · Austin · +1 more
Regulation is not optional

The EU AI Act readiness.

The EU AI Act entered into force in 2024 and its first rules, including bans on certain prohibited AI practices and an AI literacy requirement, have applied since February 2025. Rules for general-purpose AI models began applying in August 2025. The high-risk AI rules have now been pushed to December 2, 2027, while high-risk AI embedded in regulated products, such as medical devices, toys and lifts, moves to August 2, 2028.

This gives businesses more time to prepare, but it does not mean they can ignore compliance until 2027.

Gauge your readiness
Where things actually stand right now

The rollout, phase by phase.

The EU AI Act is now the world's first comprehensive AI law, and its deadlines are no longer theoretical. If your business builds, buys, or deploys AI tools, even something as simple as a customer service chatbot, some part of this law almost certainly applies to you.

Feb 202501

Bans on "unacceptable risk" uses

Social scoring, manipulation, and biometric surveillance uses are banned outright.

Aug 202502

General-purpose AI model obligations

Obligations begin for providers of general-purpose AI models.

Aug 202603

Transparency obligations apply

Disclosure requirements come into force across a broader set of AI use cases.

Dec 202704

High-risk systems deadline

Full obligations land for AI used in hiring, credit, and insurance decisions, shifted from the original August 2026 date.

Aug 202805

Embedded AI systems deadline

Obligations extend to AI embedded inside other regulated products and systems.

The four risk tiers, in plain terms

Not all AI is regulated equally.

Banned

Unacceptable risk

Social scoring, manipulative, and biometric surveillance uses, prohibited outright.

Documented

High-risk

Documentation and human oversight required: hiring, credit, and insurance decisions sit here.

Disclosed

Limited risk

Disclosure required: users must know they're interacting with AI.

Open

Minimal risk

Largely unregulated: most everyday AI tooling falls here today.

What this means practically

Start with an inventory.

Inventory every AI tool currently in use or planned, map each one against the Act's risk tiers, and prioritize documentation and disclosure for anything touching hiring, lending, or safety decisions first.

Talk to us about compliance
Maximum penalty
€35M
or 7%

of global annual turnover, whichever is higher, once each deadline lands.

This is an indicative snapshot, not a legal determination. Confirm your specific obligations under the EU AI Act with legal counsel.