The Legal Landscape of AI Interviews in 2026: NYC, Illinois, California, EU
Nothing here is legal advice. This is a candidate-framed summary of the laws that exist, what they require employers to do, and what rights they give you. Verify with an employment attorney for your specific situation.
Why These Laws Exist
AI in hiring entered widespread use before meaningful regulation caught up. Employers adopted HireVue and similar tools in the 2010s with no disclosure requirements, no bias audit obligations, and no candidate rights. The 2022-2026 regulatory wave is catching up. NYC was first with an enforceable AEDT law in 2023. Illinois’s AIVIA was first in the nation (2020). The EU AI Act classifies hiring AI as high-risk, though the obligations for that category have been pushed back to December 2027 under the EU’s Digital Omnibus package. Expect more states and countries to follow.
NYC Local Law 144 (AEDT)
Enforcement from 5 July 2023Applies to
Employers or employment agencies using an Automated Employment Decision Tool (AEDT) to evaluate candidates or employees who reside in New York City. Remote roles applicable if the candidate lives in NYC - the law follows the candidate, not the employer’s office location.
Employer obligations
- ■ Annual independent bias audit
- ■ Published audit summary on employer website
- ■ 10 business days notice before AEDT use
- ■ Disclosure of characteristics analysed
- ■ Instructions for accommodation or alternative process
Your rights as a candidate
Enforcement and penalties
DCWP (NYC Department of Consumer and Worker Protection). Civil penalties of $500-$1,500 per day per violation. 2026 enforcement phase: higher penalties expected, more frequent investigations. As of April 2026, DCWP is actively investigating employer compliance.
What you can do
Search the employer’s careers website for their published bias audit summary. The law requires it to be there. If you cannot find it, ask HR for the URL. If they cannot provide it, they may be non-compliant. File a complaint with DCWP if your rights are denied.
Illinois Artificial Intelligence Video Interview Act (AIVIA)
In force since 1 Jan 2020AIVIA was the first US state law specifically governing AI analysis of video interviews. It predates the wave of AEDT laws that followed NYC’s approach.
Applies to
Any employer using AI to analyse a candidate’s video interview for Illinois positions. Covers facial expression analysis, word choice, tone, and any AI analysis of video content.
Employer obligations
- ■ Prominent standalone disclosure (not buried in ToS)
- ■ Explanation of AI mechanics used
- ■ Disclosure of specific characteristics evaluated
- ■ Consent required before AI analysis
- ■ Records retained for 4 years
Your rights
Right to withhold consent (though the interview may still proceed for human review). Right to request video deletion. Right to notice of specific characteristics evaluated. If you did not receive a standalone notice, the employer is non-compliant with AIVIA. Enforcement via Illinois Department of Labor.
California: SB 1001 and the Broader Framework
Ongoing 2019-2026California’s regulatory environment affecting AI in hiring is a patchwork rather than a single consolidated law. Key components in 2026:
SB 1001 (Bot Disclosure, 2019)
Requires bots interacting commercially to disclose that they are not human. Narrow direct application to employment but overlaps with AI interviewing where a chatbot is used without disclosure.
CCPA / CPRA
Gives candidates data rights including access to personal data, right to know what data is collected, and limited opt-out rights. Applies to biometric data captured in video interviews.
FEHA automated decision-making guidance (2023-2026)
California Fair Employment and Housing Act regulatory guidance has been expanding to cover AI in employment decisions, though specific AEDT regulations as of April 2026 are still developing.
EU AI Act - High-Risk Hiring AI
High-risk provisions: 2 December 2027The EU AI Act classifies AI systems used in employment decisions, including screening and evaluation of candidates, as high-risk AI systems. These high-risk obligations were originally due to take effect 2 August 2026, but the EU’s Digital Omnibus package (formally adopted by the European Parliament on 16 June 2026 and the Council on 29 June 2026, awaiting publication in the Official Journal) defers them to 2 December 2027. Separately, the Act’s Article 50 transparency duty, which requires disclosure when you are interacting with an AI system, still applies from 2 August 2026.
Employer obligations (from Dec 2027)
- ■ Mandatory risk assessments
- ■ Technical documentation of the AI system
- ■ Bias testing with documented results
- ■ Human oversight of decisions
- ■ Transparency to candidates
- ■ Logging of automated decisions
Your rights (from Dec 2027)
- ✓ Right to meaningful information about automated decisions
- ✓ Right to explanation
- ✓ Right to human oversight
Penalties
Up to EUR 15M or 3% global turnover for deployer (employer) non-compliance. EUR 35M or 7% for prohibited practices.
Once the high-risk obligations take effect (now December 2027 under the Digital Omnibus deferral), EU candidates can demand explanations of automated decisions and escalate to national AI oversight bodies. In the meantime, existing GDPR automated decision rights (Article 22) already apply to AI interview scoring that constitutes solely automated decision-making, and the Act’s Article 50 transparency duty applies from August 2026.
ADA and Equality Act 2010 (Accommodation Frameworks)
These apply regardless of AI use and complement jurisdiction-specific AEDT laws.
ADA Title I (US)
Reasonable accommodation for disability applies at all hiring stages including AI screening. You describe the functional need, not the diagnosis. Employer must engage in good-faith interactive process. Cannot withdraw candidacy for requesting accommodation.
Equality Act 2010 (UK)
Reasonable adjustments duty applies to all recruitment stages. Definition of disability is broad. If the AI screening creates substantial disadvantage linked to disability, employer must take reasonable steps to avoid that disadvantage.
Comparison Table
| Jurisdiction | Notice required | Bias audit | Consent required | Alt. process | Penalties |
|---|---|---|---|---|---|
| NYC LL144 | 10 business days | Annual, independent, public | No | Yes | $500-$1,500/day |
| Illinois AIVIA | Standalone notice | Not required | Yes, before analysis | No specific requirement | Civil action |
| California (FEHA) | Emerging | Not formally required | Partial (CCPA) | Not specified | DFEH enforcement |
| EU AI Act | Yes (Dec 2027+) | Yes (technical docs) | GDPR consent rules | Human oversight right | EUR 15M or 3% |
| ADA (US) | N/A | N/A | N/A | Alternative accommodation | EEOC enforcement |
| Equality Act 2010 (UK) | N/A | N/A | N/A | Reasonable adjustment | Employment tribunal |