EU AI Act.
Is your AI compliant?
The EU AI Act is the world's first comprehensive AI regulation. It classifies AI systems by risk level — from banned practices to minimal obligations. Transparency duties apply since August 2026; high-risk obligations follow on 2 December 2027.
until High-risk AI (Annex III) · 2 December 2027
AI systems classified by risk
The EU AI Act uses a risk-based approach. All providers and deployers of AI systems in the EU market must classify their systems into one of four risk tiers.
Social Scoring
BannedBiometric Surveillance
BannedCredit Scoring
High RiskRecruitment AI
High RiskMedical AI
High RiskChatbots
Limited RiskRecommender Systems
Limited RiskSpam Filters
Minimal RiskHigh-risk AI systems: examples from Annex III
An AI system is high-risk when it is used in one of these areas (Annex III of Regulation 2024/1689), or when it is a safety component of a product covered by EU product law (Annex I), such as machinery, toys or medical devices.
- 1 Biometrics
- Remote biometric identification, emotion recognition, categorisation by sensitive traits
- 2 Critical infrastructure
- Safety components in road traffic and the supply of water, gas, heating and electricity
- 3 Education
- Admission decisions, grading, monitoring students during exams
- 4 Employment
- CV screening, targeted job ads, promotion and termination decisions, performance monitoring
- 5 Essential services
- Credit scoring, eligibility for public benefits, life and health insurance pricing, emergency call triage
- 6 Law enforcement
- Assessing the risk of becoming a victim or reoffending, evaluating evidence, profiling in investigations
- 7 Migration and border control
- Risk assessment of travellers, examining asylum, visa and residence applications
- 8 Justice and democracy
- Assisting courts in applying the law, systems meant to influence elections or voting behaviour
Not every system in these areas is high-risk: under Article 6, a provider can show that it poses no significant risk, but must document that assessment before placing the system on the market.
The cost of non-compliance
Banned AI practices
€35M
or 7% of global annual turnover
whichever is higher
High-risk violations
€15M
or 3% of global annual turnover
whichever is higher
Incorrect information
€7.5M
or 1.5% of global annual turnover
whichever is higher
What the AI Act requires — and what SiteGuardian monitors
The EU AI Act defines obligations across the AI lifecycle. SiteGuardian monitors the requirements that map to technical and compliance posture checks.
Risk Management
MonitoredSiteGuardian tracks AI-related compliance through questionnaire assessment, maps AI system risks to regulatory requirements, and monitors your overall risk posture across all applicable frameworks.
Transparency
AI system documentation and user notification requirements. Providers must ensure high-risk AI systems are designed to be sufficiently transparent for deployers to interpret output. This is an organisational measure.
Human Oversight
High-risk AI systems must be designed to allow effective human oversight, including human-in-the-loop, human-on-the-loop, or human-in-command approaches. This is an organisational measure.
Accuracy & Robustness
High-risk AI systems must achieve appropriate levels of accuracy, robustness, and cybersecurity. Technical testing and validation requirements. This is an organisational measure.
Deployer Obligations
MonitoredSiteGuardian monitors AI system performance and compliance posture, generates audit evidence for deployer obligations, and tracks whether your organisation meets its duties under the AI Act.
Transparency for Limited Risk
Chatbot and deepfake disclosure obligations. Users must be informed when interacting with AI systems. SiteGuardian covers this through questionnaire-based assessment of your AI system inventory.
Algorithmic Transparency
MonitoredSiteGuardian's DSA checks cover recommender system transparency requirements (cross-reference to Art. 27 DSA), ensuring platforms disclose their algorithmic recommendation parameters.
Reporting Obligations
MonitoredProviders and deployers must report serious AI incidents. SiteGuardian supports incident detection and reporting workflows, tracks notification deadlines, and generates pre-filled reports for authorities.
Key deadlines
The AI Act enters into force in phases. Mark these dates.
1 August 2024
Entry into force
Regulation 2024/1689 published in the Official Journal and enters into force.
2 February 2025
Banned AI practices
Prohibitions on unacceptable-risk AI systems apply: social scoring, real-time biometric surveillance, manipulative AI, emotion recognition in workplaces/schools.
2 August 2025
GPAI rules
Rules for general-purpose AI models (including foundation models) and governance structures apply.
2 August 2026
Transparency obligations
Article 50 transparency duties apply: disclosing AI interaction, marking synthetic content, labelling deepfakes. Untouched by the 2026 postponement.
2 December 2027
High-risk AI (Annex III)
Requirements for high-risk AI systems in areas such as employment, education, credit scoring and essential services. Postponed from 2 August 2026 by the AI Digital Omnibus.
2 August 2028
High-risk AI (Annex I)
Requirements for high-risk AI embedded in regulated products (medical devices, machinery, toys) under existing EU product safety legislation.
Start preparing today
Scan your website to see where you stand. SiteGuardian maps findings to AI Act articles — so you know exactly what to address.
Free forever for 1 monitor. No credit card required.