EU AI Act compliance — classification, controls and evidence on one platform
The EU AI Act applies to every AI system deployed in the EU and carries fines up to €35M or 7% of global turnover. AI-Koutsi classifies every use case in your portfolio against the Act's risk tiers, applies the right control set and generates the conformity and DPIA evidence that high-risk systems require.
Risk-tier classification for every use case
AI-Koutsi auto-classifies each use case as prohibited, high-risk, limited-risk or minimal-risk based on its intended purpose, sector and data. Reclassification is automatic when the use case changes scope.
Conformity assessment and DPIA templates
High-risk systems require conformity assessments, data governance documentation and post-market monitoring. AI-Koutsi generates the templates and keeps the artifacts versioned with the use case.
Audit trail by default
Every classification decision, control assignment and policy override is logged with the responsible role and timestamp. Auditors get the same live view as the CDO — no document hunt at audit time.
Frequently asked questions
- When does the EU AI Act apply?
- The EU AI Act entered into force in 2024 and obligations phase in through 2027. Prohibited practices apply from early 2025; high-risk system obligations apply from August 2026; general-purpose AI rules apply from August 2025. AI-Koutsi tracks the timeline per use case.
- Do we need EU AI Act compliance if we only deploy AI internally?
- Yes. The Act applies to internal AI systems too, including employment, credit scoring, critical infrastructure and many enterprise back-office use cases that fall under the high-risk tier.
- What evidence does a high-risk AI system require?
- Conformity assessment, risk management documentation, data governance records, technical documentation, logging, transparency information, human oversight design and post-market monitoring. AI-Koutsi generates the templates and keeps the evidence linked to the live use case.