AN ACT CONCERNING AUTOMATED DECISION SYSTEMS PROTECTIONS FOR EMPLOYEES.

2026 Regular Session Introduced by Saud Anwar and 32 co-sponsors

SB 435 requires Connecticut employers to disclose when they use AI in hiring and firing decisions, ensuring these automated systems are accurate and non-discriminatory while all...

FAV. RPT., TAB. FOR CAL., SEN.

#advertising #artificialintelligence #audits #civilactions #collectivebargaining #complaint #computersoftware #confidentiality #crimevictims #disclosure #discriminationinemployment #domesticviolence #electronicdata #electronicgovernmentinformation #employees #employers #employment #employmentagencies #fringebenefits #geneticinformation #hazardoussubstances #humanrightsandopportunities #commissionon #labordepartment #laborunions #notices #occupationaltraining #personaldata #publicwelfare #recordsretention #sexualassault #stateagencies #unfairordeceptivetradepractices #violations #auditors #bigotryandbias #municipalemployees #stateemployees #unemploymentcompensation

Legislative bill overview

SB 435 seeks to establish protections for employees against automated decision systems used by employers in Connecticut. The bill would regulate how employers can deploy artificial intelligence and algorithmic tools in employment decisions, including hiring, promotion, discipline, and termination. It likely requires employers to disclose when automated systems are being used, ensure systems are accurate and non-discriminatory, and provide employees with rights to appeal or challenge decisions made by these systems.

Why is this important

As AI and algorithmic decision-making become increasingly prevalent in workplace management, employees face risks of opaque, potentially biased automated decisions that affect their livelihoods. This bill addresses a growing gap in employment law by establishing baseline standards and transparency requirements. It also positions Connecticut as an early adopter of algorithmic accountability in the workplace, potentially influencing other states.

Potential points of contention

Business compliance costs: Employers may argue the requirements for testing, documentation, and appeals processes create significant administrative and financial burdens, particularly for small businesses.

Definition ambiguity: Disagreement likely exists over what constitutes an "automated decision system"—whether basic scheduling software or performance metrics qualify, which affects scope and enforcement.

Competitive concerns: Employers may resist disclosure requirements, claiming proprietary algorithms are trade secrets that provide competitive advantages.

Enforcement mechanisms: Questions about which agency enforces violations, what penalties apply, and whether private lawsuits are permitted remain critical unresolved issues.

Technological feasibility: Disputes may arise over whether bias detection in algorithms is technically feasible or sufficiently reliable for legal standards.