Connecticut's New AI Employment Law: What Employers Need to Know
By Josh Auxier
Artificial intelligence is now a routine part of the modern workplace. Employers increasingly use AI-powered tools to screen resumes, evaluate applicants, assist with performance reviews, and streamline employment decisions. Under the Connecticut Artificial Intelligence Responsibility and Transparency Act (Public Act 26-15, the “CART Act”), Connecticut employers that use certain AI technologies in employment decisions will face heightened scrutiny beginning October 1, 2026, followed by new disclosure and pre-decision notice requirements that take effect October 1, 2027.
The message from Connecticut lawmakers is clear: employers can use AI, but they cannot treat it as a black box.
New Disclosure Requirements for Employers
Connecticut's new law applies to “automated employment-related decision technology” used as a substantial factor in employment-related decisions, including hiring, promotions, discipline, termination, and other terms and conditions of employment.
Beginning October 1, 2027, before making an employment-related decision using AI, employers will be required to provide written notice to employees or applicants disclosing:
That an automated employment-related decision technology is being used;
The purpose of the technology and the employment decision being made;
The trade name of the AI tool being used;
The categories of personal data the technology will analyze and how that information will be assessed;
The sources of the personal data being used;
Contact information for the employer.
In addition, beginning October 1, 2027, employers that deploy an automated employment-related decision technology intended to interact directly with applicants or employees such as an AI-driven chatbot or interview tool, must inform those individuals, in plain language, that they are interacting with the technology, unless a reasonable person would find that fact obvious.
For many employers, this means taking a closer look at the software and AI tools already being used throughout the hiring and employment process.
AI Is Not a Defense to Discrimination
One of the most significant aspects of Connecticut's new law is what it does not allow employers to do: use AI as a defense in discrimination claims.
If an employer makes an employment decision that results in unlawful discrimination, it cannot simply argue that “the AI made the decision.” The employer remains responsible for the outcome. Connecticut law expressly provides that the use of automated employment-related decision technology is not a defense to an employment discrimination complaint.
That raises an important question for employers: How do you assess intent when an AI system is involved?
Most discrimination claims focus on the employer’s discriminatory intent against an employee or applicant. AI complicates that analysis, because employers often cannot say precisely how an algorithm weighed the information it was given or why it reached a particular recommendation. The employer cannot bury its head in the AI sand to try to avoid discrimination claims. Moreover, the result can still create a disparate impact on a group of applicants or employees, even without intent to do so.
The Risk of Unintentional Bias
Even when employers have no discriminatory intent, AI tools can create disparate impacts on protected groups, violating Connecticut's anti-discrimination laws.
Consider a simple example. An AI screening tool reviewing resumes may identify gaps in employment history as a negative factor when ranking applicants. A candidate who took time away from work after having a child could be disproportionately affected, despite being qualified for the position. The employer may have had no intention of discriminating, but the result could nevertheless have an unlawful discriminatory impact, as statistics indicate that women more often take time away from the workforce after the birth of a child (see U.S. Census Bureau, The Cost of Motherhood on Women’s Employment and Earnings (2020)).
AI systems learn from data. If the data or the criteria being evaluated reflect historical biases or unintended assumptions, those issues can become embedded in the decision-making process. For employers, that means the focus cannot simply be on whether discrimination was intended. It must also be on whether the technology produces discriminatory outcomes. This may create more disparate impact liability on the part of employers.
Practical Steps Employers Can Take Now
Employers that are considering or are already using AI tools should begin preparing now.
Some practical steps include:
Identifying all AI technologies used in employment-related decisions
Understanding what data those technologies collect and analyze
Reviewing vendor disclosures and contracts
Conducting anti-bias testing where appropriate
Regularly auditing AI-generated recommendations and results
Maintaining meaningful human oversight of employment decisions
Connecticut's new law specifically allows courts and the Commission on Human Rights and Opportunities to consider evidence of anti-bias testing and other proactive efforts designed to prevent discrimination. Documenting that work now can pay off later if a decision is ever challenged.
Employers should not assume that adopting AI automatically improves decision-making. In many cases, human review remains one of the most effective tools for identifying and correcting potential problems before they become legal issues.
Looking Ahead
AI has the potential to make workplaces more efficient and improve employment processes. But as Connecticut's new law demonstrates, increased use of AI comes with increased responsibility.
Enforcement of the CART Act’s employment provisions rests exclusively with the Connecticut Attorney General; the law does not create a private right of action. Violations are treated as unfair or deceptive trade practices under the Connecticut Unfair Trade Practices Act. For alleged violations occurring on or before December 31, 2027, the Attorney General must offer a 60-day cure period before initiating an enforcement action.
By October 1, 2026, employers using AI in hiring and employment decisions should be prepared to ensure those tools are used fairly, thoughtfully, and in compliance with the new anti-discrimination requirements. Then, by October 1, 2027, employers must also provide affected employees and applicants with disclosures explaining how those AI tools are used in employment-related decisions.
The takeaway is simple: AI can be a valuable tool, but it is not a substitute for sound judgment or compliance with Connecticut employment laws.