What regulatory frameworks do you cover?
▼
We audit for EU AI Act compliance, GDPR fairness obligations, algorithmic transparency requirements under US state laws, and platform accountability standards. Our reports are structured for regulatory submission and board attestation.
Does the API require algorithm changes?
▼
No. The API audits production algorithms without modification. We analyze decision outputs to detect bias patterns, demographic disparity, and fairness violations. Your algorithm runs unchanged; we measure and report.
What types of algorithms can be audited?
▼
Ranking, matching, recommendation, pricing, credit scoring, hiring, content moderation, and eligibility-determination systems. Any algorithm making consequential decisions about access, pricing, or eligibility is in scope.
How is audit data handled?
▼
Data stays in your infrastructure. The API runs on your servers or VPC. We don't retain algorithmic outputs, personal data, or decision histories. GDPR-compliant with full audit logs for compliance teams.
What's the typical audit timeline?
▼
Small systems: 48-72 hours. Mid-scale (1M+ decisions): 5-10 days. Complex multi-stage algorithms: 2-3 weeks. Express audits available at premium rates. Scoping call takes 30 minutes to estimate your timeline.
How often should we run audits?
▼
High-risk systems under EU AI Act require continuous monitoring. We recommend quarterly audits for systems that update frequently, annually for stable algorithms. Audit before major algorithm changes or regulatory reviews.
What does the audit report include?
▼
Fairness metrics by protected class, disparity quantification, rule impact analysis, feature contribution breakdown, remediation recommendations, and a regulatory attestation summary. Reports are tailored for technical teams and boards.
Is API-level integration required?
▼
Most flexible: API integration for real-time audits. Batch mode for historical datasets. SQL-based for data warehouse analysis. We support your infrastructure; no lock-in or specialized tooling required.
What if bias is found?
▼
We identify the root cause - feature engineering, training data, or decision thresholds. The report specifies which demographic groups are disadvantaged and by how much. We recommend remediation strategies; you own the fix timeline.
How do we get started?
▼
Email compliance@audit.example.com or book a scoping call at our website. We'll assess your system, timeline, and regulatory context. Formal proposal and contract within 48 hours.