AI in Employment: State Compliance Updates for Employers

As artificial intelligence continues to transform the modern workplace, state legislatures are moving quickly to establish guardrails around the use of automated tools in employment decisions. Recent developments in Colorado, Connecticut, and Illinois reflect a broader trend toward increased oversight of AI and automated decision-making systems used in hiring, promotion, compensation, and other employment-related decisions.

For HR professionals, staying ahead of these developments is essential. While AI tools may help streamline recruiting and workforce management processes, they can also create compliance risks if they are not reviewed, monitored, and used consistently with applicable anti-discrimination, privacy, and notice requirements.

Below is a summary of recent developments in Colorado, Connecticut, and Illinois, including what employers should know and practical steps HR teams can take to prepare. While these laws address similar technologies, their terminology, scope, and compliance requirements differ by state, so employers should avoid assuming that the same tools, decisions, or workflows are covered in each jurisdiction.

Colorado Repeals and Reenacts AI Protections with Key Changes

As an update to our prior post, Colorado recently passed Senate Bill 26-189, which replaces the artificial intelligence provisions adopted under the state’s 2024 AI framework with a revised compliance structure for automated decision-making technology, or ADMT. While the prior law established one of the nation’s first broad state AI regulatory frameworks for the deployment of AI, the new law significantly reshapes Colorado’s approach to AI regulation. Under this framework, accountability requirements apply when ADMT is used to make consequential decisions, including decisions related to employment, hiring, compensation, promotion, scheduling, performance management, discipline, access to benefits, or selection for training and development opportunities.

Employers that use covered ADMT in employment-related decisions should review the law’s updated notice and disclosure requirements, documentation and recordkeeping obligations, and procedures for responding to requests for meaningful human review and reconsideration following an adverse decision. Many of these provisions are scheduled to take effect January 1, 2027.

Key provisions include:

• Effective Date and Rulemaking Requirements: Beginning January 1, 2027, employers that use covered ADMT in employment-related consequential decisions should be prepared to request, review, and retain developer or vendor documentation addressing the tool’s intended uses, categories of training data, known limitations, and instructions for appropriate use and human review. The Colorado Attorney General is also directed to adopt rules by January 1, 2027, clarifying certain post-adverse outcome disclosure requirements.
• Notice Requirements: Employers using covered ADMT must provide clear and conspicuous notice at the point of interaction. If the ADMT results in an adverse outcome, the employer must provide a plain-language description of the ADMT’s role within 30 days after the consequential decision.
• Human Review: Individuals have the right to request meaningful human review and reconsideration after a covered ADMT makes a consequential decision that results in an adverse outcome.

Connecticut Expands Oversight of AI and Automated Employment Decisions

Connecticut has enacted Public Act No. 26-15, which includes new requirements for automated employment-related decision technology, or AEDT. For employers, AEDT generally includes systems that process personal data and produce outputs, such as predictions, scores, rankings, classifications, or recommendations, which are a substantial factor in making or materially influencing employment decisions such as hiring, promotion, discipline, discharge, renewal of employment, and selection for training or apprenticeship.

Effective October 1, 2026, the law clarifies that using an automated system is not a defense to a discrimination claim, although bias testing and risk-mitigation efforts may be considered. The law also creates disclosure obligations when AEDT is used with applicants or employees, including notice when an individual is interacting with AEDT and written notice before AEDT is used to support an employment decision.

Illinois Moves Forward with Proposed AI Notice Rules

Illinois’ AI employment law took effect on January 1, 2026, requiring employers to notify applicants and employees when AI is used in employment decisions and prohibiting the use of AI in a way that results in discrimination based on protected classes. The law also prohibits employers from using ZIP codes as a proxy for protected classes.

Since our prior update, the Illinois Department of Human Rights has moved forward with proposed rules addressing the law’s notice requirements, including when notice must be provided, how notice must be delivered, and when the notice obligations apply. The public comment period remains open through June 29, 2026, so employers should continue to monitor the rulemaking process while reviewing their current AI use for alignment with the law’s existing requirements.

Recommended Next Steps for Employers

Employers using, or considering the use of, AI or automated decision-making tools in HR processes should consider taking the following steps:

• Identify AI Tools. Identify all AI-driven or automated tools currently used in recruiting, hiring, promotion, compensation, performance management, discipline, discharge, layoffs, or other employment-related decisions.
• Review Vendor Documentation Agreements. Confirm whether vendors can provide documentation explaining how their tools operate, what data they use, known limitations, whether bias testing is performed, and how human review may be incorporated.
• Assess Notice Obligations. Determine whether applicable state laws require notice to applicants or employees when AI or automated decision-making tools are used, including whether notice must be provided before, during, or after a decision.
• Maintain Human Oversight. Ensure that AI-assisted employment decisions are monitored by a human, based on job-related criteria, and do not create or reinforce discriminatory outcomes.
• Track Data. Bias testing requires demographic data about impacted applicants and employee pre- and post-employment decision. Determine what data to gather and how frequently to measure for bias.
• Develop a Multi-State Compliance Approach. Multi-state employers should map AI tools use by jurisdiction and determine which state-specific notice, documentation, human review, anti-discrimination, or recordkeeping requirements apply. Employers may want to adopt a uniform baseline process, with state-specific workflows where needed.

How HR Works Can Help

HR Works provides industry expertise and guidance to HR practitioners navigating the evolving legislation of artificial intelligence. Our governance tools include bias testing, inventory tracking, disclosure templates, vendor vetting protocols, and legislative updates, within a suite of other resources designed to outfit HR with the information needed to keep up with this rapidly developing field.

Contact us to learn more about how HR Works can support your AI compliance efforts.