Make talent quality your leading analytic with skills-based hiring solution.
ID verify compliance now means something noticeably more specific than it did even two years ago. Employment eligibility rules like E Verify are only one layer. A newer, fast growing body of state law now directly regulates the AI tools, biometric checks, and video interview technology recruiters use to verify and evaluate candidates in the first place, and several of these laws carry real financial penalties.
Treating ID verify compliance as settled once E Verify and standard background checks are in place is no longer accurate. The newer AI and biometric specific rules are where most current compliance risk actually sits, and that landscape has shifted meaningfully even within 2026, which is exactly why this guide is dated and will need revisiting again as rulemaking continues.
Local Law 144 prohibits using an automated employment decision tool for NYC hiring or promotion decisions unless the employer has completed an independent annual bias audit and published the results publicly for at least six months. Employers must give candidates advance notice at least 10 business days before using such a tool, and candidates must be offered an alternative process where one exists.
Penalties start at 500 dollars for a first violation and range from 500 to 1,500 dollars for each subsequent violation, and since each day of continued unlawful use, and each missed candidate notice, can count as a separate violation, exposure can accumulate quickly across a high volume hiring process.
Illinois has regulated AI analyzed video interviews since 2020 under the AI Video Interview Act, and those requirements remain directly relevant to any recruiter using video screening today: employers must give applicants notice before the interview that AI will analyze the video and assess fitness for the position, obtain written consent, and explain in plain terms what the AI evaluates. Video sharing is restricted, applicants can request deletion, and employers must comply within 30 days. Where AI solely determines which applicants advance, annual demographic reporting is also required.
As of January 1, 2026, a second and considerably broader law also applies: House Bill 3773, which amends the Illinois Human Rights Act. Where the 2020 law covered video interviews specifically, HB 3773 covers any AI that influences or facilitates an employment decision, recruitment, promotion, discipline, or termination included, not just interviews. It requires employers to notify applicants and employees when AI is used, naming the AI product and its developer, which employment decisions it affects, what personal data it processes, a point of contact for questions, and information about reasonable accommodations, delivered annually to current employees, within 30 days of adopting a new AI system, and to applicants through job postings.
This section corrects a claim on the previous version of this page, and the correction matters enough to call out directly. Colorado’s original, broad Artificial Intelligence Act (CAIA), enacted in 2024, was expected to take effect in mid 2026 with requirements including annual impact assessments and a NIST aligned risk management program. It did not take effect on that timeline. Following sustained industry opposition and a federal legal challenge, Colorado delayed the law repeatedly through the first half of 2026, then repealed and replaced it entirely.
The replacement, the Automated Decision Making Technology Act (ADMTA), takes effect January 1, 2027, pending completion of rulemaking by the Colorado attorney general, and is substantially narrower than the original CAIA. It drops the impact assessment and NIST risk management requirements and focuses on three things for employers using AI in consequential employment decisions: notice to the candidate or employee before a covered automated tool is used, a plain language explanation of the decision and the tool’s role within 30 days of an adverse outcome, with a designated, trained person available to conduct a genuine human review, and three year record retention. The ADMTA is enforced solely by the Colorado attorney general and does not create a private right of action, unlike New York City’s law.
Any recruiting team that has been tracking the original Colorado AI Act’s June 2026 timeline should update that internally. The obligations that actually apply, and when, changed twice this year.
Maryland’s HB 1202 requires written consent before an employer uses facial recognition technology on a job applicant, a narrower but still binding requirement relevant to any biometric verification step. More states, including Washington, New Jersey, California, and Texas, currently have similar bias audit and disclosure legislation pending, and Colorado’s experience in 2026 is a useful reminder that pending legislation can change substantially, or slip its timeline, before it actually takes effect, which is one more reason to track effective dates rather than headlines alone.
This growing legal backdrop is exactly why the identity verification case made in our companion piece on candidate identity checks has shifted from optional to expected. Compliance risk is now a second, independent driver alongside fraud prevention.
Disclose AI and biometric tool usage clearly and in advance rather than in fine print discovered later, since nearly every law summarized above centers on advance notice and consent as the baseline requirement. Build a genuine human review path into any AI scored step, and document it, since both Colorado’s new ADMTA and NYC’s Local Law 144 anchor their requirements around a demonstrable human oversight layer.
Treat biometric verification and any identity verification step as a specific compliance surface with its own consent and disclosure requirements, not as a general purpose fraud control that automatically satisfies every legal obligation around it. Our identity verification FAQ covers the most common questions recruiting teams raise when building this into an existing process.
Track effective dates, not just headlines. As Colorado’s 2026 timeline shows, a law that is widely reported as taking effect on a given date can be delayed, amended, or replaced before it actually applies, and building compliance around the wrong date creates real exposure.
ID verify compliance in 2026 and heading into 2027 is a moving target rather than a fixed checklist, with new state laws layering on top of established employment eligibility rules faster than most recruiting teams can track informally, and with at least one major law, Colorado’s, changing shape twice within a single year.
Building disclosure, consent, and human review into the process by default, rather than retrofitting it after a specific state’s law takes effect, or citing a timeline that has since moved, is the more durable approach as this landscape keeps expanding. Glider’s identity verification tools are built around exactly this disclosure and human review pattern.
ID verify compliance refers to meeting the growing set of federal and state legal requirements governing identity verification, AI hiring tools, and biometric technology used during recruiting, including disclosure, consent, bias auditing, and human review obligations.
NYC Local Law 144 prohibits using an automated employment decision tool for hiring or promotion in New York City without an independent annual bias audit, public disclosure of results, and at least 10 business days advance notice to candidates, with penalties starting at 500 dollars per violation and rising to 1,500 dollars for repeat violations.
AI proctoring and AI analyzed video interviews are legal in most jurisdictions but increasingly regulated, requiring advance notice, written consent, and clear disclosure of what is being evaluated under laws such as Illinois’s AI Video Interview Act and HB 3773, and New York City’s Local Law 144.
Effective January 1, 2026, HB 3773 requires employers to notify applicants and employees whenever AI influences an employment decision, naming the AI tool and its developer, the decisions it affects, the personal data it processes, and accommodation information, delivered annually and whenever a new AI system is adopted.
No. The original Colorado AI Act was repealed before taking effect. Its replacement, the Automated Decision Making Technology Act, takes effect January 1, 2027, and is narrower in scope, focused on notice, human review after an adverse decision, and record retention rather than mandatory impact assessments.
Penalties vary by jurisdiction. NYC Local Law 144 penalties start at 500 dollars per violation and rise to 1,500 dollars for repeat violations. Colorado’s ADMTA is enforced by the state attorney general without a fixed per violation fee or a private right of action, while Illinois’s laws tie noncompliance to demographic reporting and consent obligations rather than a flat penalty schedule.
Yes. Maryland’s HB 1202 requires written consent from a job applicant before an employer uses facial recognition technology during the interview process, a narrower requirement focused specifically on biometric identification rather than AI scoring generally.
Second device cheating is when a candidate uses a phone, tablet, or second monitor outside the webcam’s view during a proctored online assessment to look up answers, message someone for help, or mirror their screen to a helper. It is one of the hardest forms of assessment fraud to stop with software alone, because the […]
Contingent worker identity verification is the practice of confirming that a freelancer, contractor, or gig worker is who they claim to be before they are onboarded, not just screened on paper. Most companies apply this scrutiny to full time employees and skip or shortcut it for contingent talent, which is exactly the gap fraud actors […]
A structured interview scorecard is a standardized form that lists the specific competencies a role requires, gives every interviewer the same rating scale to judge candidate answers against, and forces a written justification for each score. It exists to replace gut feeling with evidence, so five interviewers evaluating the same candidate land on comparable, defensible […]
