When AI screens your application, what you can actually ask for

Software has been ranking applications for years. What changed recently is that a handful of governments decided candidates should be told about it — and, in a few places, be able to do something about it.

The rights are real but uneven. They depend on where you are, where the employer is, and which rule happens to apply. Enforcement is thinner than the headlines suggest. This is a map of what actually exists as of August 2026, and what it is worth doing with it.

This is general information, not legal advice. If something here matters to a decision you are making, check the primary sources listed at the end or talk to a lawyer.

New York City: notice, and a published bias audit you can read

New York City's Local Law 144 covers "automated employment decision tools" used on candidates in the city. Employers using one must commission an independent bias audit every year, publish a summary of the results publicly, and tell candidates before the tool is used.

Two details are worth knowing because they are useful to you specifically.

The notice has to arrive at least 10 business days before the tool is used, and it has to say what the tool assesses — the job qualifications and characteristics it is scoring you on, not just that "AI is involved."

The notice must also include how to request an alternative selection process or an accommodation. Read that carefully: the law requires the employer to tell you how to ask. It does not, on its own, guarantee the answer is yes. That is still better than not knowing the option exists.

The bias audit summary is the part almost nobody uses. It is supposed to be public on the employer's site, and it reports adverse impact across demographic groups. If you are deciding whether to spend a weekend on an application, a company that publishes a real audit and a company that publishes nothing are telling you different things about how they hire.

Illinois: they need your consent before AI reads your video interview

Illinois has had the Artificial Intelligence Video Interview Act since January 1, 2020, and it is narrower but sharper. It applies specifically to AI analysis of recorded video interviews.

Before the interview, the employer must tell you AI may be used to analyze it, explain how the AI works and what general types of characteristics it uses to evaluate applicants, and obtain your consent. The statute is blunt about the consequence: an employer may not use AI to evaluate applicants who have not consented.

Consent to the AI analysis is a separate thing from agreeing to record a video. Those are not the same permission.

Two further rights sit inside the same act. The employer may not share your video except with people whose expertise or technology is needed to evaluate you. And if you ask them to delete it, they have 30 days to delete it and to instruct anyone who received a copy to delete theirs, including backups.

Colorado, from 2027: a route back to a human

Colorado is the one to watch rather than to use today. The state's original 2024 AI Act was replaced: Senate Bill 26-189 was signed on May 14, 2026, repealing and reenacting the earlier provisions with new requirements, and it takes effect January 1, 2027.

For someone being screened, the interesting parts are what it gives you after a decision goes against you. A deployer must give notice at the point of interaction, and following an adverse outcome must provide a plain-language description of the tool's role within 30 days. You can request correction of your personal data. And you can request meaningful human review and reconsideration of a consequential decision that produced an adverse outcome.

That last one is the right that most of the current US rules do not give you. It does not exist yet — January 2027 — so treat it as a reason to keep an eye on Colorado postings, not as something to invoke now.

Europe: the deadline you may have read about has moved

If you saw that EU rules on hiring AI arrive in August 2026, that was true when it was written and is no longer.

The AI Act does classify recruitment and candidate-selection systems as high risk. But the Digital Omnibus on AI, Regulation (EU) 2026/1744, was published in the Official Journal on July 24, 2026 and entered into force on July 27 — days before the original deadline. It defers the high-risk obligations for standalone Annex III systems, which is the category recruitment tools fall into, to December 2, 2027.

Some transparency obligations stayed on the original schedule. The substantive high-risk duties that would most affect how you are screened did not.

The part nobody advertises: almost nobody is enforcing this

New York State's Comptroller audited how the city enforces Local Law 144 and published the results in December 2025. The findings are worth reading before you assume the rules protect you by default.

The audit covered July 2023 to June 2025. Over those two years the city's Department of Consumer and Worker Protection received two complaints about automated employment decision tools, and the audit identified no enforcement actions taken — against a penalty regime that runs to $1,500 per day.

The department does not proactively monitor compliance. It relies on educating stakeholders and waiting for complaints. It ran one survey of 32 companies' websites and found a single compliance issue; an independent review of those same 32 companies identified at least 17 instances of potential non-compliance.

The gap between one and seventeen is the whole story. The rules are on the books. Whether they are followed in your particular application is, in practice, something you find out by asking.

What to actually do

None of this is worth much as trivia. It is worth something as a small set of habits.

• Before you invest real time in an application, look for the bias audit summary on the employer's careers or jobs page if you are applying in New York City. Its presence or absence tells you something.
• If you receive a notice that an automated tool will be used, read what it says the tool assesses. That is a description of the scoring criteria, handed to you in writing.
• Ask, in writing, whether an automated tool is used in screening and what it evaluates. A written question creates a record and is a reasonable thing for a candidate to ask.
• Where an alternative selection process or accommodation is offered, request it if a tool is likely to read you badly for reasons unrelated to the job.
• For a recorded video interview in Illinois, remember consent to AI analysis is specific and separate. You can also ask for the video to be deleted afterwards.
• Keep your own copy of what you submitted and when. If you ever need to contest a decision, that record is the thing you will wish you had.

What this does not mean

It does not mean an automated screen is illegal, or that being filtered by one is evidence of anything improper. Most of these tools do work employers would otherwise do by hand, at volume, and the rules above mostly govern disclosure rather than outcome.

It also does not mean you should write your application for a machine. The criteria a tool scores are usually the criteria in the posting. Answering the posting clearly and specifically is the same work either way.

What has changed is that in some jurisdictions you are entitled to know a tool is there, to know what it claims to measure, and — soon, in Colorado — to ask a person to look again. Those are worth using while they exist.

Sources

Every consequential claim above comes from one of these. Where a rule is recent or contested, check it yourself before relying on it.

• New York City Administrative Code, Local Law 144 of 2021, Automated Employment Decision Tools — NYC Department of Consumer and Worker Protection: https://www.nyc.gov/site/dca/about/automated-employment-decision-tools.page
• Office of the New York State Comptroller, "Enforcement of Local Law 144 — Automated Employment Decision Tools," published December 2, 2025, covering July 2023 to June 2025: https://www.osc.ny.gov/state-agencies/audits/2025/12/02/enforcement-local-law-144-automated-employment-decision-tools
• Illinois Artificial Intelligence Video Interview Act, 820 ILCS 42, effective January 1, 2020 — Illinois General Assembly: https://www.ilga.gov/Legislation/ILCS/Articles?ActID=4015&ChapterID=68
• Colorado Senate Bill 26-189, Automated Decision-Making Technology, signed May 14, 2026, effective January 1, 2027 — Colorado General Assembly: https://leg.colorado.gov/bills/sb26-189
• Regulation (EU) 2026/1744, the Digital Omnibus on AI, published in the Official Journal July 24, 2026, deferring Annex III high-risk obligations to December 2, 2027. The regulation amends Regulation (EU) 2024/1689, the AI Act.

Current as of August 2026. Rules in this area are changing quickly and several dates above have already moved once.