Article 4
AI literacy
Train staff who use, supervise, or manage AI systems.
Due: August 2025
← All regulations
EU and national authoritiesAugust 2026Regulation (EU) 2024/1689
EU AI Act
The EU AI Act classifies AI systems by risk. Organisations using AI need an inventory, AI literacy, transparency processes, and stronger controls for high-risk use cases.
Who may be in scope
Indicative practical impact for EU SMBs.
- Companies using AI for HR, finance, scoring, safety, customer support, or internal decision support.
- SaaS teams providing AI features to customers in the EU.
- Organisations using generative AI in operational processes.
- Providers and deployers of high-risk AI systems listed in Annex III.
Key obligations
What often needs evidence or review.
Article 26
AI use inventory
Record purpose, owner, risk, input data, and business process for each AI system.
Due: August 2026
Article 14
Human oversight
Assign responsibility for decisions supported by high-risk AI systems.
Due: August 2026
Article 50
Transparency
Inform users when they interact with AI or receive AI-generated content where required.
Due: depends on use
Fines and sanctions
Risk is measured in money, contracts, and lost trust.
Violation typeMaximum sanctionEnforcement
Prohibited AI practicesEUR 35 million or 7% of worldwide turnoverNational authority
High-risk AI obligationsEUR 15 million or 3% of worldwide turnoverNational authority
Incorrect information to authoritiesEUR 7.5 million or 1% of worldwide turnoverNational authority
How Splnit.eu helps
Turn obligations into controls, evidence, and deadlines.
AI inventory
Track tools, owners, purposes, risk levels, and review status.
AI policy
Generate working policies for acceptable AI use and human oversight.
AI literacy evidence
Record training completion and staff acknowledgement.