EU AI Act Compliance Scorecard
Illustrative example of ICOSA's multi-model assessment output against the EU AI Act (2024/1689) — representative verdict patterns across common architecture classes, not a completed audit of named third-party systems.
Illustrative Example Data
The systems below are labeled by architecture class, not by real named vendors or products — this table demonstrates ICOSA's assessment output format and typical verdict distribution, not results from an actual completed scan of specific third-party AI systems. This does not constitute legal advice or an official regulatory determination.
| System | Provider | Verdict | Confidence | Risk Flags |
|---|---|---|---|---|
| Constitutional-Aligned Model | Architecture Class 2 | Compliant | 94% | None |
| Frontier Transformer A | Architecture Class 1 | Compliant | 92% | Transparency |
| EU-Hosted MoE Model | Architecture Class 4 | Compliant | 91% | None |
| Multimodal Transformer B | Architecture Class 3 | Compliant | 89% | Data Governance |
| RAG-Native Model | Architecture Class 5 | Compliant | 88% | Documentation |
| Compact Model A | Architecture Class 15 | Compliant | 88% | None |
| Open-Weight Model A | Architecture Class 6 | Compliant | 87% | Oversight |
| Hybrid Architecture Model | Architecture Class 9 | Compliant | 86% | None |
| Compact Model B | Architecture Class 16 | Compliant | 86% | Documentation |
| Multilingual Model | Architecture Class 7 | Compliant | 85% | TransparencyData Governance |
| Conversational Model | Architecture Class 11 | Compliant | 84% | Documentation |
| Adversarial-Tuned Model | Architecture Class 8 | Compliant | 83% | Documentation |
| RAG-Native Model B | Architecture Class 19 | Compliant | 82% | Documentation |
| MoE Model B | Architecture Class 10 | Non-Compliant | 78% | TransparencyOversightData Governance |
| MoE Model C | Architecture Class 14 | Non-Compliant | 74% | DocumentationTransparency |
| Open-Weight Model B | Architecture Class 12 | Non-Compliant | 72% | TransparencyRisk Management |
| Large-Scale Model A | Architecture Class 18 | Non-Compliant | 71% | Data GovernanceTransparency |
| Open-Weight Model C | Architecture Class 13 | Non-Compliant | 69% | TransparencyData GovernanceOversight |
| Open-Weight Model D | Architecture Class 17 | Non-Compliant | 65% | Risk ManagementTransparencyOversight |
Methodology
Each AI system is assessed by the ICOSA multi-model council using Byzantine Fault Tolerant consensus. The council evaluates compliance across all relevant articles of the EU AI Act, including transparency obligations (Art. 50, 52), risk management (Art. 9), data governance (Art. 10), human oversight (Art. 14), and documentation requirements (Art. 11, 12).
Confidence scores represent the degree of consensus among council members. Higher confidence indicates stronger agreement on the verdict. Risk flags identify specific areas where deficiencies were detected during assessment.
Assessments use the Sentinel (3-model) tier for initial screening, with Baseline (5-model) and Full Certification (11-model) available for deeper analysis upon request.
Get Your System Scanned
EU AI Act obligations are already phasing in — get your AI system assessed now, rather than waiting for the next deadline to catch you unprepared.
Start with Sentinel Assessment