概述
Classification follows intended purpose and actual use. The Annex III high-risk obligations for these use cases apply from 2 December 2027 under the AI Omnibus.
深入探讨
Hiring software is not automatically high-risk just because it uses automation. Annex III covers systems intended for recruitment or selection, including targeted job advertisements, analysis and filtering of applications, and evaluation of candidates. It also covers systems used to make decisions affecting work-related terms, promotion, termination, task allocation based on individual behavior or personal traits, and monitoring or evaluation of workers. The purpose and deployment context matter: a general scheduling tool may be different from a system that scores workers for consequential decisions. Where a system falls in the high-risk category, the Act divides responsibilities. Providers must meet the applicable system requirements, including risk management, data governance, technical documentation, logging, transparency, human oversight, accuracy, robustness, and cybersecurity. Deployers must use it in accordance with instructions, assign competent oversight, monitor operation, and keep relevant logs under their control. Their use of output must also comply with employment, equality, privacy, and other applicable law. The AI Act does not itself authorize an employer to make a particular decision. The AI Omnibus, which entered into force on 27 July 2026, sets 2 December 2027 for Annex III high-risk obligations. The prohibition on emotion recognition in workplaces already applies from 2 February 2025, except where the Act’s medical or safety exception applies. This prohibition is separate from the high-risk classification: a prohibited practice is not made acceptable by adding a human reviewer or documenting it as high-risk. Before procurement, an employer should identify the exact task, affected workers, decision authority, vendor role, and any downstream action. Ask the provider for intended purpose, known limits, input requirements, and instructions. Test for disparate errors using lawful, representative evidence, define when a person can override or stop use, and provide a channel for correction or contest. Do not treat a model score as a neutral fact about a person.
战略影响
风险与安全
灾难性和日常的人工智能危害都取决于谁了解风险以及谁能够采取行动。
更清晰的判决
公众和专业素养决定强有力的安全政策在政治上是否可行。
打破炒作
清晰的解释可以减少炒作、实验室公关和模糊道德剧场的影响。
The Future of EU AI Act Rules for AI in Hiring and HR
The EU Commission’s current implementation timeline places Annex III employment and worker-management requirements on 2 December 2027. Guidance, standards, and national enforcement practice may clarify how providers and employers demonstrate compliance. Organizations can prepare by inventorying systems and decisions now, documenting vendor roles, and testing governance in real workflows. Any date-sensitive policy should be checked against the consolidated Regulation and Commission updates because later amendments can change obligations or exceptions. Keep dated records of the applicable text, guidance, and decisions so teams can explain their reasoning when rules or system purposes change.
现实世界的实施
A recruiter inventories a CV parser that filters applicants and asks the vendor for its stated purpose and limitations.
An employer separates a shift-planning tool from a worker-scoring feature that recommends disciplinary review.
A company stops workplace emotion inference unless a documented medical or safety exception truly applies.
A human reviewer records reasons for overriding an AI shortlist and can pause the workflow when errors appear.
风险与防护栏
将存在风险视为科幻小说，同时能力复合。
混淆了表面产品安全与高度自治下的对准。
只给非英语和非专业观众留下低质量的资源。
实施路线图
单独的产品危害、误用和失控/失调风险。
询问哪些证据会改变您对时间表和严重性的看法。
比起营销主张，更喜欢主要来源和具体评估。
确定一条行动路径：职业、政策、资金或技能——而不仅仅是意识。
不断探索
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the EU AI Act Rules for AI in Hiring and HR quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
常见问题
What is EU AI Act Rules for AI in Hiring and HR?
The EU AI Act lists certain employment and worker-management systems as high-risk, including tools used to recruit, select, evaluate, promote, or allocate tasks based on personal traits. Classification follows intended purpose and actual use. The Annex III high-risk obligations for these use cases apply from 2 December 2027 under the AI Omnibus.
Which use is named in Annex III employment examples?
Recruitment filtering and candidate evaluation are listed high-risk use cases.
Which practice is already prohibited in workplaces, subject to a narrow exception?
The Act prohibits workplace emotion recognition with stated exceptions.
Does adding human review make a prohibited practice acceptable?
A prohibited practice remains prohibited unless an exception applies.
For a deployer of a high-risk system, which step is required?
The Act sets deployer duties including oversight and monitoring.
Why should a hiring team measure false negatives?
Error types can reveal exclusion risks hidden by aggregate accuracy.
继续学习
相关指南
为此主题精选的更多指南