IS AI MONITORING OR EVALUATING WORKER PERFORMANCE OR BEHAVIOUR?
Annex III lists AI intended to monitor and evaluate the performance and behaviour of persons in work-related contractual relationships as high-risk.
WHEN AI WATCHES OR SCORES WORKERS,
THE EMPLOYMENT DECISION HAS TO STAY REVIEWABLE.
Workplace AI can influence promotion, termination, task allocation, performance assessment and day-to-day worker treatment. Annex III treats specified worker-management and monitoring uses as high-risk, while Article 5 separately prohibits workplace emotion inference except for medical or safety reasons.
BOUNDARYThis page is educational and operational guidance. It is not legal advice, certification, conformity assessment or regulatory approval.
CURRENT SOURCE ANCHOR
OPEN OFFICIAL EU SOURCE ↗EUR-Lex · Regulation (EU) 2024/1689 · Article 5 and Annex III, Employment
Annex III lists specified AI used to affect employment terms, promotion, termination, task allocation, and monitoring or evaluation of worker performance and behaviour as high-risk. Article 5 prohibits workplace emotion inference except for medical or safety reasons. Current Commission guidance places Annex III employment high-risk rules on the 2 December 2027 timeline.
WHAT BUSINESSES NEED TO ESTABLISH
Do not start with a generic checklist.
Start with the actual system, role, use case and evidence boundary. The same regulation can produce different obligations for different actors and systems.
IS AI ALLOCATING TASKS OR AFFECTING PROMOTION, TERMINATION OR WORK TERMS?
AI used to allocate tasks based on individual behaviour or personal traits, or to make decisions affecting work-related terms, promotion or termination, also sits on the Annex III employment high-risk route.
IS THE SYSTEM INFERRING WORKER EMOTIONS?
Article 5 prohibits AI systems used to infer emotions of natural persons in the workplace, except where the system is intended for medical or safety reasons. A prohibited use cannot be fixed merely by adding high-risk controls.
CAN YOU PROVE THE HUMAN DECISION AND WORKER-IMPACT PATH?
Preserve the system and model version, monitored signals, scoring or evaluation output, decision rules, human review, overrides, notices, complaints or challenges, logs, limitations, and the evidence supporting any consequential employment action.
THE TA-14 OPERATING ROUTE
Turn the question into a governed record.
The goal is not merely to reach an answer. It is to preserve what facts, evidence, scope and limitations supported that answer at that time.
Define the workplace function precisely
Separate productivity analytics, scheduling, task allocation, safety monitoring, performance scoring, conduct monitoring, promotion support, termination support and purely administrative tools. Different functions can create different legal routes.
Test prohibited practices before high-risk classification
If the system infers emotions from biometric data in the workplace, test the Article 5 prohibition first. Do not treat prohibited emotion inference as a normal Annex III deployment unless a valid medical or safety exception applies.
Bind human oversight and worker-impact evidence to the decision
Connect monitored data, model output, human reviewer, employment authority, override, notice, challenge or grievance path, logs, performance and bias evidence, and the resulting work-related decision.
Revalidate after model, policy, metric or workflow changes
A new scoring model, changed performance metric, broader monitoring scope, new sensor source, removed human checkpoint, changed task-allocation logic or expanded employment purpose can make the previous classification and evidence stale.
READY TO OPERATE · START PAID ACCESS NOW
Move from reading about the EU AI Act to maintaining the record.
Choose the smallest operating tier that fits the portfolio today. Upgrade when system count, team size or governance scope actually requires it.
Evidence Passport
$19/MOKeep a living system-level evidence record with obligations, gaps, versions and revalidation state.
START EVIDENCE PASSPORT →Compliance Workspace
$49/MOCoordinate evidence, owners, documentation, incidents and team compliance work in one governed workspace.
START COMPLIANCE WORKSPACE →Governance Pro
$99/MOOperate broader high-risk, GPAI, FRIA, post-market and material-change governance across a growing portfolio.
START GOVERNANCE PRO →Institution
$499/MORun institutional governance with expanded users, authority workflows, examiner rooms and portfolio reporting.
START INSTITUTION →Need an independent human readiness examination?A governed readiness review is separate from software access and starts at $750.
REQUEST READINESS REVIEW →Not ready to subscribe? Classify first.
Use the free classifier to establish the system, intended purpose, possible actor role, EU exposure and unresolved facts. When continuing evidence infrastructure is needed, paid access begins at $19 per month. Independent human readiness review remains a separate service.
COMMON QUESTIONS
Questions businesses are asking now.
Is employee-monitoring AI high-risk under the EU AI Act?
Certain systems are. Annex III includes AI intended to monitor and evaluate the performance and behaviour of persons in work-related contractual relationships, as well as specified systems used for task allocation and decisions affecting employment terms, promotion or termination.
Is all workplace analytics automatically high-risk?
No. The classification depends on intended purpose and whether the system falls within an Annex III use case or another high-risk route. Purely administrative or non-consequential tools may fall outside that category, but the actual function must be assessed.
Can employers use AI to infer worker emotions?
Article 5 prohibits workplace emotion inference, except where the system is intended for medical or safety reasons. That prohibition has applied since 2 February 2025.
When do the Annex III employment high-risk rules apply?
Current European Commission guidance states that high-risk rules for Annex III employment systems apply from 2 December 2027.
What evidence should be preserved for workplace AI?
Preserve system identity and version, intended purpose, provider and deployer roles, monitored inputs, scoring or evaluation outputs, human review, overrides, notices, logs, performance and bias evidence, complaints or challenges, material changes and revalidation history.
Does TA-14 certify employee-monitoring or workplace AI as compliant?
No. TA-14 can preserve the classification basis, prohibited-practice analysis, authority, evidence, worker-impact record, human-oversight state, gaps and revalidation history. It does not provide legal advice, certification, conformity assessment or regulatory approval.