Employers that use software to sort job applicants now face disclosure rules in at least two places. Since January 1, 2026, Ontario requires publicly advertised job postings from employers with 25 or more employees to state whether artificial intelligence is used to screen, assess, or select applicants. In New York City, Local Law 144 bars an employer from using an automated employment decision tool unless the tool had an independent bias audit within the past year and candidates received notice.
Key takeaways
- Ontario's rule covers postings made on or after January 1, 2026, and a general statement that AI is used is enough.
- New York City requires an annual independent bias audit, a public summary of the results, and notice to candidates 10 business days before use.
- Neither rule bans automated screening or sets a pass mark: the New York City law does not require any action based on audit results.
- Statistics Canada measured 12.2 percent of businesses using AI in the second quarter of 2025, but that figure covers all uses, not hiring.
- Colorado enacted a new AI law for consequential decisions, employment included, on May 14, 2026, with some duties starting January 1, 2027.
What Ontario requires in a job posting
The Ontario government's guide to the Employment Standards Act says a publicly advertised posting must disclose AI use if the employer uses it to screen, assess, or select applicants. The guide describes AI as a machine-based system that infers from its inputs to generate outputs such as predictions, content, recommendations, or decisions, and says whether a given system qualifies depends on the facts. Its wording on detail is brief: it is enough to state that AI is used for those purposes.
The duty stays with the employer when a recruiting firm does the screening. Employers with fewer than 25 employees on the day of posting are outside the rule, as are the Crown and certain Crown agencies. Employees are counted individually, so part-time and casual staff each count as one. Employers must keep a copy of each posting, with any application form, for three years after public access ends. The guide separates preliminary screening against minimum qualifications from an interview that assesses suitability, but it does not define "screen" further for AI tools.
What New York City requires
The city's Department of Consumer and Worker Protection (DCWP) enforces Local Law 144 of 2021, and its FAQ on automated employment decision tools sets out the mechanics. A covered tool uses machine learning, statistical modeling, data analytics, or AI, helps make employment decisions, and substantially assists or replaces discretionary decision-making. Screening counts as an employment decision. The law applies to jobs tied to an office in the city, including fully remote jobs associated with one, and enforcement began on July 5, 2023.
A bias audit is an impartial evaluation by an independent auditor. At a minimum it calculates selection or scoring rates and impact ratios across sex, race and ethnicity, and intersectional categories. Auditors cannot work for the employer or the tool's vendor, and a category under two percent of the data may be left out. Employers must publish a summary of the latest audit and the date they began using the tool, and they must send notice, including how to request a reasonable accommodation, 10 business days before use. The vendor is not responsible for the audit. The employer is.
The two rules compared
The table sets the main features side by side. It is a simplified orientation based on the two government sources named below, and an employer's counsel decides how the rules apply to a specific hiring process.
| Feature | Ontario (ESA job-posting rules) | New York City (Local Law 144) |
|---|---|---|
| Start | Postings made on or after January 1, 2026 | Law effective January 1, 2023; enforcement from July 5, 2023 |
| Who is covered | Employers with 25 or more employees | Employers and employment agencies using a covered tool for jobs tied to a city office |
| Main duty | Say in the posting that AI is used to screen, assess, or select | Independent bias audit within one year, public summary, candidate notice |
| Detail required | A general statement is enough | Selection or scoring rates and impact ratios by category |
| Result of the check | Not applicable | No specific action required by the law itself |
| Discrimination claims | Handled under other employment law | Referred by DCWP to the NYC Commission on Human Rights |
What the rules leave open
Both regimes are about disclosure, not approval. Ontario does not ask what the tool measures, how it was trained, or whether a person reviews its output. New York City asks for an audit but, in DCWP's words, does not require any specific action based on the results, and discrimination complaints go to a different agency. Federal, state, and city anti-discrimination law keeps applying to the hiring decision whatever the audit says.
Some uses fall outside the New York City rule altogether. DCWP says scanning a resume bank, outreach to potential candidates, and inviting applications are not covered, because the duties attach to assessing people who applied for a specific position.
Adoption data and a new Colorado law
Official adoption statistics are broad. In the Canadian Survey on Business Conditions, Statistics Canada found that 12.2 percent of businesses had used AI to produce goods or deliver services over the previous 12 months in the second quarter of 2025, up from 6.1 percent in the second quarter of 2024. The analysis page reviewed for this article reports no hiring-specific figure, so it cannot show how many employers screen applicants this way. Adoption is also a measure of use, not of effect on jobs.
In the United States, Colorado's General Assembly lists SB 26-189 as signed on May 14, 2026. The bill page names employment among the areas that count as consequential decisions. It requires developers to give deployers documentation from January 1, 2027, and directs the attorney general to adopt rules on disclosures after an adverse outcome by that date.
Questions to ask about a screening tool
- Does the tool rank, score, or filter applicants, or does it only schedule and route them?
- Which stage of hiring does it touch: outreach, screening, interview, or promotion?
- Where is a posting, notice, or audit summary published, and who keeps the copies?
- Does the employer or the vendor hold the audit data, and is the auditor independent under the local definition?
- Can a candidate ask for an accommodation or a human review, and where is that explained?
This article gives general information about public rules. It is not legal advice, and it does not recommend a hiring product.
FAQ
Does Ontario require employers to name the AI tool they use?
No. The government guide says it is enough to state that artificial intelligence is used to screen, assess, or select applicants. The statement must appear in the posting.
Does New York City ban AI hiring tools that show bias in an audit?
The law itself does not. DCWP says it requires the audit and notices but no specific action on the results. Anti-discrimination laws still apply, and discrimination claims are referred to the city's Commission on Human Rights.
Who is responsible for the bias audit, the employer or the vendor?
The employer. DCWP states that the vendor is not responsible for the audit, although a vendor may commission one from an independent auditor and the employer remains ultimately responsible.
Do small Ontario employers have to disclose AI use?
Not under these posting rules. They do not apply to employers with fewer than 25 employees on the day the posting goes up.
Related reading
- Whether AI will replace accountants
- AI and human intelligence compared
- Could AI surpass human intelligence
- AI in warehouse management
More in AI and Society.
Last verified: October 2026
Sources
- Requirements related to publicly advertised job postings, Your guide to the Employment Standards Act, Government of Ontario, 2026
- Automated Employment Decision Tools: Frequently Asked Questions, NYC Department of Consumer and Worker Protection, June 29, 2023
- Automated Employment Decision Tools (AEDT) page, NYC Department of Consumer and Worker Protection, 2026
- Analysis in Brief on artificial intelligence use by businesses in Canada, Statistics Canada, second quarter of 2025 (published June 16, 2025)
- SB26-189 bill page, Colorado General Assembly, 2026