Connecticut Enacts SB 5: Comprehensive Workplace AI Regulation and Novel AI-Caused RIF Disclosures

Connecticut Enacts SB 5: Comprehensive Workplace AI Regulation and Novel AI-Caused RIF Disclosures

On May 27, 2026, Connecticut Governor Ned Lamont signed Senate Bill (SB) 5 (designated as Public Act No. 26-15) into law, establishing a landmark bipartisan framework for artificial intelligence (AI) regulation. While the 74-page bill covers subscription-based AI, frontier model whistleblower protections, and AI companion safety, its most immediate and heavy-hitting impact on enterprise risk teams lies in its sweeping regulation of "automated employment-related decision technology" (AERDP).

Starting on a staggered basis (with key employment provisions taking effect on October 1, 2027), the law represents one of the nation's most stringent and prescriptive workplace AI frameworks, alongside California's CCPA regulations and Colorado's rewritten AI Act (Colorado’s AI Reset: SB 26-189 Repeals and Replaces SB 24-205 Following xAI Lawsuit and DOJ Intervention).

Scope and Definition of Workplace AI (AERDP)

The law applies to "automated employment-related decision technology" (AERDP), defined as:

"any technology that processes personal data and uses computation to generate any output, including, but not limited to, any prediction, recommendation, classification, ranking, score or other information, that is a substantial factor used to make or materially influence an employment-related decision" — Ogletree Deakins

An "employment-related decision" includes hiring, promotion, discipline, discharge, tenure, terms, privileges, or conditions of employment. Common software tools (spreadsheets, word processors, spellcheckers, map navigation) are excluded, as are scheduling and productivity monitoring tools, provided they do not make material employment decisions.

Strict Employer Notice and Disclosure Mandates

Starting October 1, 2027, employers in Connecticut must provide plain-language disclosures to employees and applicants interacting with AERDP. Additionally, before making any employment decision where the automated technology is a "substantial factor" (defined as assisting in making and capable of altering the outcome), the employer must provide a written notice disclosing:

• That the technology has been deployed and its trade name.
• The purpose and nature of the employment decision.
• The categories of personal data processed and how they are assessed.
• The sources of that personal data.
• Contact information for the employer.

If an adverse employment-related decision is made, the employer must provide the affected individual with a high-level explanation of the principal reasons, the specific role of the AERDP, the data used, and an opportunity to correct their personal data.

No Defense Against Discrimination Claims

Crucially, SB 5 amends Connecticut's employment discrimination law to explicitly strip employers of an easy defense:

"SB 5 amends Connecticut’s employment discrimination law to specify that the use of covered automated employment-related decision technology to make an employment decision is 'not a defense against a complaint alleging a discriminatory practice.'¹" — Ogletree Deakins

However, proactive "anti-bias testing" or similar efforts can be considered by courts and the Connecticut Commission on Human Rights and Opportunities (CHRO) as a mitigating factor.

Novel AI-Related RIF Disclosures (Effective October 1, 2026)

In a first-of-its-kind statutory requirement, starting October 1, 2026, employers serving written notice to the Labor Department of a plant closing or mass layoff under the federal Worker Adjustment and Retraining Notification (WARN) Act must explicitly disclose:

"whether the reductions covered by the notice 'are related to the employer’s use of artificial intelligence or another technological change.'" — Ogletree Deakins

Developer-Deployer Division of Labor

To support employer compliance, the law mandates that AI developers provide deployers (employers) with all information necessary to fulfill their disclosure obligations.²

Whistleblower Protections for Frontier Developers

Starting January 1, 2027, "frontier developers" (entities developing "foundation models" capable of influencing physical or virtual environments) must establish anonymous internal reporting channels for public safety or catastrophic risks. Retaliation against whistleblowers is strictly prohibited, with civil penalties of up to $1,000 per violation.

Enforcement

There is no private right of action for violations of the notice and disclosure provisions (which are enforced exclusively by the state attorney general as unfair or deceptive trade practices, subject to a 60-day cure period for violations occurring before December 31, 2027). However, traditional discrimination claims arising from AI hiring bias can still be pursued privately under existing employment laws.