SentinelEU AI Act
⚠ SAMPLE·AUD-2026-00026·2026-07-17·HIGH_RISK — Annex III §4

TalentMatch AI v1.0

[Provider name withheld for this public sample] · HR / Recruitment & employment decisions

24files
316signals
6high findings
66/100dossier readiness
RSA-PSS · SHA-256
0/100

Compliance Score

20/100
FAILDossier readiness: 66/100 (blocked)ARTICLE_WEIGHTED_V3

Risk classification

HIGH RISK — Annex III §4

This system is classified as HIGH_RISK under the EU AI Act because it makes automated employment decisions affecting natural persons in the EU market. The full 22-article assessment applies.

Makes employment decisionsAnnex III §4
Affects natural personsArt. 6
Automated scoring componentAnnex III §4
Deployed in EU marketArt. 2
GPAI model insideArt. 51
Biometric identificationAnnex III §1

Verdict scale — where this system lands

≥ 85

Statically Aligned

≥ 65

Aligned

≥ 40

Gap

< 40

Fail

THIS SYSTEM

Article breakdown

22 articles assessed

Compliant
Aligned
Gap
Fail
ArticleRequirementScoreCoverageFindingsVerdict
Art. 14
Human OversightCritical
0/11
1fail
Art. 13
TransparencyCritical
3/10
1fail
Art. 9
Risk ManagementCritical
0/8
1fail
Art. 10
Data Governance
0/8
1fail
Art. 20
Logging & Traceability
3/8
1fail
Art. 50
AI Output Transparency
0/8
—fail
Art. 5
Prohibited Practices
7/7
—compliant
Art. 15
Robustness & Security
0/6
—fail
Art. 55
GPAI Systemic Risk Management
—
—
—n/a
Art. 47
EU Declaration of ConformityCritical
0/4
—fail
Art. 72
Post-Market Monitoring
0/4
—fail
Art. 51
GPAI Classification
—
—
—n/a
Art. 53
GPAI Transparency Obligations
—
—
—n/a
Art. 11
Technical Documentation
2/3
—aligned
Art. 12
Record Keeping
0/3
—fail
Art. 19
Automatically Generated Logs
3/3
—compliant
Art. 17
Quality Management System
0/3
—fail
Art. 27
Fundamental Rights Impact AssessmentCritical
0/3
—fail
Art. 52
GPAI Systemic Risk Reclassification
—
—
—n/a
Art. 4
AI Literacy
0/2
—fail
Art. 16
Provider Obligations
0/2
—fail
Art. 49
EU Database RegistrationCritical
0/2
—fail

Findings & remediation

Critical & high-priority findings

6 high-severity policy violations of 316 total findings (169 low · 137 info) — click any finding to see the remediation path

310 additional findings (169 low · 137 info · 4 other) included in the full audit bundle

Technical dossier

Annex IV document verification

22/22 required documents generated · 15 contain provider-input placeholders that must be completed · submission currently blocked

7 auto-generated, complete15 need provider input
Executive Summaryexecutive_summary.md—Verified
Article Coverage Detailarticle-coverage-detail.md—Verified
Risk Registerrisk_register.mdArt. 9Incomplete
Data Governancedata_governance.mdArt. 10Incomplete
Human Oversight Evidencehuman-oversight.mdArt. 14Incomplete
Instructions for Useifu.mdArt. 13Incomplete
Fundamental Rights Impact Assessmentfria.mdArt. 27Incomplete
Quality Management Systemqms.mdArt. 17Incomplete
EU Declaration of Conformityeu_doc.mdArt. 47Incomplete

Output bundle

7 outputs generated — every audit

Core
audit.json

Full machine-readable report — RFC8785-Lite canonical

Core
report.html

Standalone visual report, no external dependencies

DevOps
SARIF v2.1.0

GitHub Security tab · VS Code · any CI/CD pipeline

DevOps
SBOM CycloneDX

AI component inventory for supply chain compliance

Legal
Annex IV pack

Pre-filled Annex IV technical documentation templates

Security
checksums.txt

SHA-256 hashes of all output files

Security
RSA-PSS signature

Cryptographic proof — verifiable without Sentinel

Audit serial

AUD-2026-00026

Signature

RSA-PSS · SHA-256 · RFC8785-Lite

Engine

v1.3.1 · ai-act-rules-v1.3.1

Model

ARTICLE_WEIGHTED_V3
Run this on your actual system

Your system. Your audit. In minutes.

Connect your repository and get the exact same format — applied to your actual codebase. Signed, reproducible, legally formatted. Ready for a Notified Body.

Start free audit How it works
Zero code executionRSA-PSS signed22 articles scoredAnnex IV generatedSARIF for CI/CDReproducible output