Prepare for new automated hiring and pay-transparency rules

Automated screening tools and pay-transparency mandates are moving from “nice to have” to “must have” across the U.S. and Europe. Between bias-audit duties for automated hiring systems, expanding disclosure rules in job ads, and new limits on salary-history questions, 2026 is shaping up to be a major compliance inflection point.

For HR, Talent Acquisition, Legal, and Compensation teams, the challenge is not just tracking new laws, it’s operationalizing them: mapping where tools are used, validating vendor claims, publishing the right notices, and producing pay ranges that reflect reality. Below is a practical guide to preparing for new automated hiring and pay-transparency rules.

1) Automated hiring rules are no longer theoretical, NYC’s AEDT regime is already here

New York City’s “Automated Employment Decision Tools” (AEDTs) framework under Local Law 144 continues to shape how employers must deploy algorithmic tools in hiring and employment decisions well into 2026. NYC prohibits employers and employment agencies from using an AEDT unless it has had a bias audit within the last year, the required audit information is made publicly available, and required notices are provided to candidates and employees.

This is not just about “AI” in the abstract. If your process uses software to screen resumes, rank applicants, score video interviews, or recommend candidates, and that output meaningfully influences decisions, your workflow may fall within the AEDT concept and trigger audit-and-notice duties. That means compliance requires coordination between TA operations, HRIS/ATS owners, procurement, and legal.

Actionably, organizations should maintain an inventory of hiring and internal mobility tools, document where an automated score or recommendation is used, and ensure bias audits are current (within one year). You’ll also need a durable method to publish audit summaries and to deliver candidate/employee notices consistently across roles, business units, and recruiting channels.

2) Expect more scrutiny even when enforcement is uneven

Even if enforcement feels inconsistent, scrutiny is rising. In December 2025, a New York State Comptroller audit concluded that NYC’s Department of Consumer and Worker Protection (DCWP) is “falling short in enforcing Local Law 144” regulating AI in employment decisions. That kind of finding can trigger reforms, increased enforcement focus, or policy updates, often quickly.

Separately, NYC enacted Local Law 2026/025 directing the city to conduct a study and issue a public report on the impacts of algorithmic tools (including AEDTs) on employees. A city-backed report can influence future rulemaking, enforcement priorities, and the expectations of applicants, employees, labor groups, and the media.

The practical implication: don’t treat today’s enforcement climate as your long-term risk baseline. Build a program that would hold up under renewed audits, press questions, or regulator inquiries, especially around documenting tool purpose, governance, audit recency, and how notices are delivered and retained.

3) Colorado’s AI Act delay is a window to prepare, not a reason to wait

Colorado’s AI Act (SB 24-205), expected to affect AI used in hiring and other “high-risk” contexts including employment, had its effective date delayed by 2025 amendments from February 1, 2026 to June 30, 2026. Many organizations will interpret the delay as extra runway, and it is, but the work is substantial.

High-risk AI compliance typically requires more than policy updates. It can involve risk assessments, governance controls, vendor oversight, incident processes, and documentation that demonstrates reasonable care. For employers, this can touch recruiting, promotion, scheduling, performance management, and other worker-impacting systems depending on how the law is interpreted and applied.

Use the delay to conduct a gap assessment: identify where algorithmic tools are used, evaluate whether they meaningfully influence employment outcomes, and establish a review process for new tools before deployment. If you rely on vendors, require clear documentation on model purpose, training data considerations (where available), testing, and how bias and errors are monitored over time.

4) The EU AI Act makes employment AI “high-risk”, and deadlines are coming fast

In the EU AI Act, AI used for employment and worker management is categorized as “high-risk.” According to EU Commission materials, transparency rules apply in August 2026, and high-risk obligations roll out in August 2026 and August 2027. Employer-oriented explainers also note that high-risk system requirements generally take effect on August 2, 2026, with guidance expected by February 2, 2026.

For employers operating in or hiring into the EU, this means you should plan for a world where certain employment AI tools come with mandatory risk controls, documentation, and transparency expectations. Even if a vendor “owns” the model, employers still need to understand what the tool does, when it is used, and how it affects candidates and workers.

A practical preparation step is to map EU-touching workflows: sourcing, screening, assessments, interview scheduling tools, performance analytics, and internal mobility systems. Then align procurement requirements with the EU AI Act timeline, ensuring vendors can support documentation, transparency, and any required controls before the August 2026 milestone.

5) Pay transparency is becoming standard operating procedure across U.S. states

Pay-transparency rules now differ widely by jurisdiction, but the direction is consistent: more disclosure, earlier in the process, and more specificity. Minnesota requires that, since January 1, 2025, employers with 30+ employees may not post a job without including the starting salary range (or a fixed pay rate) and a general description of benefits.

New Jersey guidance describes a pay and benefits transparency law effective June 1, 2025 for employers with 10+ employees. Job postings must include an hourly wage/salary or range and a general description of other compensation programs for which the employee would be eligible, pushing employers to describe not only base pay, but broader compensation structures.

Vermont’s Act 155 (signed June 4, 2024) takes effect July 1, 2025 and requires most employers with 5+ employees to include compensation or a “good faith” range expected for the position in job ads. Massachusetts requirements reportedly took effect October 29, 2025 for employers with 25+ employees, including ranges in postings and disclosures for promotions/transfers, expanding transparency beyond external recruiting.

6) “Pay range” quality matters, California tightens definitions and Washington debates fixes

Publishing a range is not enough if the range is not credible. California’s SB 642, effective January 1, 2026, tightens the “pay scale” definition so that ranges must reflect a good-faith estimate of what the employer reasonably expects to pay upon hire. This pushes organizations away from overly broad “cover everything” ranges and toward ranges that align to actual hiring intent.

Washington has also been actively refining its approach. A 2025 amendment proposal discussed allowing employers a short cure window (commonly described as 5 or 14 days) after notice to correct deficient postings before penalties or damages, and excluding certain third-party replicated postings. Even without final details, the debate shows regulators are paying attention to posting accuracy and enforcement mechanics.

Operationally, this is where compensation governance meets recruiting execution. If recruiters, hiring managers, and external agencies are posting roles quickly, you need a controlled source of truth for ranges and benefits language, plus a way to audit postings for compliance. Many employers are implementing posting templates, pre-approved comp language, and automated checks in ATS-to-career-site publishing flows.

7) The EU Pay Transparency Directive forces earlier disclosure and blocks salary-history questions

In Europe, the EU Pay Transparency Directive sets a national implementation deadline of June 7, 2026. Directive summaries emphasize two particularly disruptive changes: candidates must receive initial pay level or pay range in the job ad or before the first interview, and employers may not ask candidates about salary history.

For organizations accustomed to negotiating pay late in the process, or using salary history as a benchmark, this requires redesigning the front end of recruiting. It also raises the stakes on job architecture (levels, families, locations) and on how ranges are set for new hires versus internal transfers.

Preparation steps include standardizing when and how pay ranges are communicated, training recruiters on what they can and cannot ask, and updating interview guides and application forms to remove salary-history prompts. It’s also wise to reconcile pay ranges used in postings with internal pay equity audits so that transparency does not surface unexplained inconsistencies.

8) Plan for cross-border complexity, and the wild card of federal preemption debates

Multi-jurisdiction compliance is becoming the norm: a single job posting might be visible in multiple states, and a single assessment tool might be used for candidates in NYC, Colorado, and the EU. Add to that Delaware’s pay transparency law signed in September 2025, with enforcement beginning September 26, 2027 and requiring a good-faith minimum-to-maximum pay range plus benefits description, and it’s clear that requirements will keep layering.

There is also policy uncertainty. In 2025, reporting described a debate over a proposed U.S. budget provision that would block enforcement of state AI laws for 10 years. Whether or not such a provision ever becomes law, the discussion itself is a reminder to build compliance programs that are resilient to legal changes, modular controls, good documentation, and the ability to pivot quickly.

The best strategy is to build a single internal standard that meets (or closely approaches) the strictest requirements you face, then localize where needed. For automated hiring, that often means adopting audit-ready documentation and transparent candidate communications. For pay transparency, it often means consistent range-setting methodology, posting templates that include benefits language, and a clear policy on when ranges can differ by location.

Preparing for new automated hiring and pay-transparency rules is less about one-off legal updates and more about building repeatable systems. Employers should treat AEDT and AI governance like any other enterprise risk: define ownership, document tool use, validate vendors, and ensure notices and audits are not ad hoc.

On the pay side, the organizations that will adapt best are those that can explain their ranges, keep postings consistent, and train teams to communicate pay earlier and more confidently. With major milestones arriving through 2026, and more likely to follow, now is the time to turn compliance into a durable operating capability.