Home / Compliance guides / EU AI Act
COMPLIANCE GUIDE · EU AI ACT
EU AI Act guide for US and Canadian companies
Why the EU AI Act reaches companies outside Europe, how its risk levels work, the key deadlines and what providers and deployers of AI must do.
Updated November 2026
Regulation (EU) 2024/1689
10-minute read
THE SHORT ANSWER
The EU AI Act (Regulation (EU) 2024/1689) is the first comprehensive AI law. It applies to companies anywhere that place AI systems on the EU market or whose AI output is used in the EU. It sorts AI by risk: some practices are banned, high-risk systems face strict duties and some systems carry transparency rules. Obligations phase in from February 2025 to August 2027.
WHO NEEDS IT
It applies outside the EU too
If EU customers use your AI, or its output is used in the EU, the Act can apply even if you have no office there.
01
AI providers selling into the EU
SaaS and AI vendors with EU customers are providers under the Act.
02
Deployers using AI in the EU
Companies using AI for hiring, credit or other decisions about people in the EU.
03
General-purpose AI model providers
Developers of foundation models have their own documentation and transparency duties.
KEY REQUIREMENTS
Risk levels and what they mean
§
Prohibited practices
Banned since 2 February 2025, such as social scoring and manipulative AI.
§
High-risk AI
Risk management, data governance, documentation, human oversight, logging and conformity assessment.
§
Transparency duties
Tell people when they interact with AI and label AI-generated or deepfake content.
Summary only, not legal advice. EU proposals may change some deadlines; confirm with your counsel.
HOW TO GET COMPLIANT
Getting ready in four steps
1 · AI inventory
List the AI systems you build or use, and where.
2 · Classify risk
Map each system to a risk level and your role.
3 · Close gaps
Risk management, documentation, oversight and transparency.
4 · Govern
Run an AI management system, often aligned with ISO 42001.
HOW ZUROTH HELPS
Services and proof for EU AI Act
AI Governance & ISO 42001 →
AI Strategy Consulting & Readiness Assessment →
GDPR & Data Privacy Compliance →
Generative AI & LLM Application Development →
AI Agent Development & Workflow Automation →
MLOps & LLMOps Services →
SOC 2 Compliance & Audit Readiness →
Technology Strategy & Roadmapping →
FINTECH
AI strategy
AI governance
AI strategy, portfolio and governance for regional banks
A value-scored AI portfolio, an operating model and EU AI Act-ready governance.
1
risk register for every AI system
Read the story →
MEDIA
Generative AI
Data engineering
GenAI archive search with rights built in for broadcasters
Search decades of video and audio by meaning, with rights checked before every reuse.
3
teams on one platform: newsroom, rights, archive
Read the story →
FINTECH
AI agents
DORA
Agentic AI KYC, real-time fraud scoring and DORA for payments fintechs
Faster onboarding, fewer fraud losses and regulator-ready resilience.
3
workstreams: KYC agents, fraud, DORA
Read the story →
Does the EU AI Act apply to US and Canadian companies?
Yes, in many cases. It applies to providers that place AI systems or general-purpose AI models on the EU market, and to providers and deployers outside the EU when their AI output is used in the EU. Selling AI software to EU customers usually brings you into scope.
When do EU AI Act obligations apply?
The Act entered into force on 1 August 2024. Bans and AI literacy duties apply from 2 February 2025, general-purpose AI model rules from 2 August 2025 and most other rules, including many high-risk duties, from 2 August 2026, with some product-related rules in 2027. EU proposals may move some dates.
What counts as high-risk AI?
High-risk AI includes safety components of regulated products and systems listed in Annex III, such as AI for hiring and worker management, credit scoring, education, access to essential services, law enforcement and border control. These must meet strict requirements before being placed on the market.
What are the penalties under the EU AI Act?
Fines can reach €35 million or 7% of worldwide annual turnover for prohibited practices, €15 million or 3% for most other breaches, and €7.5 million or 1% for supplying incorrect information, with lower caps for small and medium-sized companies.
How should a company start preparing?
Start with an inventory of every AI system you build or use, then classify each by risk and by your role as provider or deployer. Close high-risk and transparency gaps first, train staff on AI literacy and set up ongoing governance, often using ISO 42001.
OTHER COMPLIANCE GUIDES
Need help with EU AI Act? Let's talk.
Book a free 30-minute call with a EU AI Act specialist. You'll leave with a scope, a timeline and a fixed-price quote.
A clear scope and timeline
A fixed-price quote
No obligation, NDA on request