EU AI Act Compliance

Compliance Statement

AAMOS is committed to compliance with the EU AI Act (Regulation EU 2024/1689), which entered into force in August 2024 with phased applicability through 2027. This page sets out the risk classification of each AAMOS engine, our conformity assessment status, and our commitments to technical documentation and transparency.

Our EU compliance activities are managed by Landvex AB, Sweden (org.nr 559141-7042), as our EU legal entity. Regulatory enquiries should be directed to legal@landvex.com.

Engine Risk Classification

We have conducted a risk classification of all 12 AAMOS engines in accordance with EU AI Act Articles 6–7 and Annex III.

Engine	Risk Classification	Rationale	Status
AMOS Identity	High Risk	Biometric identification — Annex III, Point 1(a)	Conformity Assessment Underway
AMOS Fraud	High Risk	Creditworthiness and fraud scoring — Annex III, Point 5(b)	Conformity Assessment Underway
AMOS Vision	Limited Risk	General object detection; transparency obligations apply	Compliant
AMOS Infrastructure	Limited Risk	Infrastructure condition monitoring; no direct individual effect	Compliant
AMOS Safety	Limited Risk	Workplace safety monitoring; transparency obligations apply	Compliant
AMOS Inspection	Minimal Risk	Product quality control; no individual profiling	Compliant
AMOS Compliance	Limited Risk	Regulatory compliance checks; transparency obligations apply	Compliant
AMOS Reality Engine	Minimal Risk	Scene verification; no individual decision-making	Compliant
AMOS Change Engine	Minimal Risk	Temporal change detection; no individual profiling	Compliant
AMOS Risk Engine	Limited Risk	Risk scoring for assets; transparency obligations apply when used in decisions affecting individuals	Compliant
AMOS Evidence Engine	Minimal Risk	Audit chain generation; evidentiary tool without decision-making	Compliant
AMOS Prediction Engine	Limited Risk	Predictive maintenance; transparency obligations apply	Compliant

High-Risk AI System Obligations

For AMOS Identity and AMOS Fraud, classified as high-risk AI systems under Annex III, we are committed to:

Establishing and maintaining a quality management system (Art. 9)
Maintaining comprehensive technical documentation (Art. 11 and Annex IV)
Implementing record-keeping and logging capabilities (Art. 12)
Ensuring transparency and provision of information to users (Art. 13)
Designing for human oversight (Art. 14)
Achieving accuracy, robustness, and cybersecurity standards (Art. 15)
Completing conformity assessment prior to EU market placement (Art. 43)
Registering in the EU AI Act database (Art. 71) prior to deployment

Technical Documentation Commitment

Model Cards

Model cards for all engines document intended purpose, known limitations, performance metrics, and testing datasets.

Bias Assessments

Regular bias and fairness assessments are conducted for all engines, with particular rigour for Identity and Fraud.

Accuracy Metrics

Performance benchmarks are maintained and updated with each model release. Metrics are available to enterprise customers on request.

Human Oversight

All high-risk engines include API response fields that support human review workflows. We provide documentation on how to implement meaningful human oversight.

Contact

EU AI Act and regulatory compliance enquiries: legal@landvex.com

EU representative: Landvex AB, Sverige