EU AI Act Compliance

OmegaEngine aligns with obligations under the EU AI Act (2025 text), including GPAI requirements, deployer responsibilities, and high-risk AI controls. OmegaEngine acts primarily as a judgment-layer safety component and does not perform biometric categorization, social scoring, or prohibited practices.

1. Classification

• • OmegaEngine = GPAI component (general-purpose)
• • Customers = deployers
• • Integrates safely into high-risk categories by design:
•   - HR, credit, lending
•   - Healthcare triage decisions
•   - Access to essential services
•   - Law-enforcement (non-biometric)

2. Required Controls Implemented

• • Logging & traceability (Art. 12)
• • Technical documentation & model cards (Art. 13)
• • Post-market monitoring hooks (Art. 61)
• • Transparency obligations (Art. 52)
• • Data governance: minimization + lawful basis (Art. 10)
• • Human oversight assistive features (Art. 14)
• • Risk management: structured scoring (Art. 9)

3. Deployers using OmegaEngine

Customers integrating OmegaEngine into high-risk systems must:

• • Provide human oversight for BLOCK/REVIEW actions
• • Maintain internal documentation of use-cases
• • Ensure lawful basis for data inputs
• • Configure their own retention windows

4. Prohibited AI Practices

OmegaEngine **does not** and cannot be configured to:

• • Perform biometric categorization
• • Run social scoring systems
• • Manipulate vulnerable groups
• • Emotion inference in workplace/schools
• • Predict criminality or protected attributes

5. Documentation & Evidence

OmegaEngine auto-generates:

• • Input/output logs
• • Policy version applied
• • Risk score metadata
• • Time-stamped decision receipts