veille Built for, HR & Recruitment
Vertical

You screen candidates with AI. Three regulators screen you.

Hiring is the single most-regulated use of AI. The moment a tool screens, scores, or ranks an applicant, Ontario's Bill 149, the EU AI Act, and Québec's Law 25 all apply at once, job-posting disclosure, an algorithmic impact assessment, human oversight, and the right to a human review. Veille maps each obligation to each hiring tool, continuously, whether you built it or bought it.

Primary frameworkBill 149, EU AI Act Annex III, Law 25
Applies toAny AI used to screen, rank, or assess candidates
CoverageContinuous monitoring, bilingual evidence

Regulatory framework

Three frameworks converge on one use case: AI in hiring.

Ontario, Employment Bill 149, AI hiring disclosure in job postings (Ontario)
Ontario's Working for Workers Four Act, 2024 (Bill 149) added s. 8.4 of the Employment Standards Act: employers that use AI to screen, assess, or select applicants must disclose that use in publicly advertised job postings. This obligation took effect January 1, 2026. Veille encodes it and tracks the implementation details as they are clarified. Posting-level disclosure is the starting line, not the finish, the EU AI Act and Law 25 layer additional duties on the same tools.
EU, High-risk EU AI Act, Recruitment is named high-risk (Annex III)
The EU AI Act lists AI used for recruitment, filtering applications, and evaluating candidates as high-risk under Annex III, point 4. That triggers the full high-risk regime: a risk-management system (Art. 9), data governance (Art. 10), human oversight (Art. 14), accuracy and robustness (Art. 15), technical documentation (Annex IV), and a conformity assessment with an EU declaration (Art. 43). It is extraterritorial, if you hire, or could hire, anyone in the EU, you may be in scope, regardless of where your company sits. High-risk (Annex III) provisions apply 2 December 2027, deferred from 2 August 2026 by the Digital Omnibus on AI. This is the same use case other jurisdictions are racing to regulate (NYC Local Law 144, Colorado, Illinois), the trend is one direction.
Québec, Privacy Law 25, Automated hiring decisions and the right to human review (QC)
When an AI tool makes or materially assists a hiring decision about a Québec candidate, Art. 12.1 requires you to inform them that an automated system was used, the principal factors that led to the decision, and their right to have it reviewed by a human. Profiling candidates triggers additional transparency and opt-out obligations, and a Privacy Impact Assessment (EFVP / PIA) is required before the screening system goes live. In force since September 2023, with active CAI enforcement.

Use cases

Hiring AI Veille monitors, built or bought.

System Primary frameworks Key obligations
Resume / CV screener Bill 149EU AI Act Annex IIILaw 25 Applicant disclosure, Conformity assessment, Right to human review
Candidate ranking / scoring EU AI Act Annex IIIBill 149 Algorithmic impact assessment, Bias testing, Job-posting disclosure
Video / asynchronous interview analysis EU AI Act Annex IIILaw 25 High-risk classification, Human oversight, Accuracy & bias documentation
Chatbot pre-screening Law 25Bill 149 Disclosure of AI interaction, Transparency notice, Job-posting disclosure
Skills / personality assessment AI EU AI Act Annex IIIBill 149 Impact assessment, Protected-ground bias testing, Human-rights review
Sourcing / matching algorithm Law 25EU AI Act Profiling transparency, Opt-out, Provider-vs-deployer mapping

Pain points

What we hear from talent and legal teams.

01
Hiring AI is bought, not built, and no one tells legal
Recruiters adopt screening and matching tools directly from vendors. Each one is an AI system with disclosure, registry, and assessment obligations the moment it touches a candidate. Veille discovers these tools continuously, including the ones procurement never logged.
02
"We disclose it in the posting" isn't enough
Disclosure is the easy obligation. Bill 149 makes the posting-level disclosure mandatory in Ontario; the EU AI Act then adds a conformity assessment, human oversight, and bias documentation; Law 25 adds a transparency notice and a right to human review. A single line in the job ad does not satisfy the rest. Veille maps the full obligation set to each tool.
03
A rejected candidate's complaint is now a regulatory event
Law 25 gives candidates a right to a human review of an automated decision; human-rights law gives them a discrimination claim. Without documented oversight and bias testing, a single complaint becomes an inquiry. Veille keeps the evidence ready before the question is asked.
04
"The vendor says they're compliant", but the obligation is yours
As the deployer, you carry the disclosure, oversight, registry, and assessment obligations even when a third party built the model. Veille separates provider from deployer duties and tracks the ones that land on you.

Act now

Law 25 has been in force since 2023. Ontario's AI-hiring rules and the EU AI Act are next.

A 30-minute call is enough to map which of your hiring tools are in scope, which obligations apply, and where the gaps are. No deck. No sales process.

Book a 30-minute call →