Connecticut Attorney General Maps Existing Laws to AI Use
Connecticut Attorney General William Tong issued an advisory this week explaining how the state's existing civil rights, privacy, data security, competition and consumer protection laws apply to artificial intelligence systems. The guidance serves as an enforcement roadmap for businesses operating in Connecticut and a rights reminder for residents.
The advisory signals the AG's enforcement priorities even as Connecticut lacks a dedicated AI regulation law. Tong said businesses cannot treat AI deployment as a regulatory blank slate.
Civil Rights and Antidiscrimination
The advisory identifies hiring, employment, healthcare, housing, insurance and lending as contexts where AI systems can trigger civil rights violations. Connecticut's antidiscrimination laws apply regardless of whether a system is federal or state-regulated.
The Trump administration's focus on AI deregulation does not suspend federal antidiscrimination laws, which the Connecticut AG can enforce.
Privacy and Data Use for AI Training
Connecticut's Data Privacy Act (CTDPA) governs personal data fed into AI systems. Businesses must comply with data minimization, protection assessments, notice and consent requirements.
A June 2025 amendment to the CTDPA adds a specific requirement effective July 1, 2026: companies must disclose in their privacy notices whether they "collect, use or sells personal data for the purpose of training large language models."
The advisory clarifies that businesses buying datasets from third-party data brokers must verify those brokers provided proper notice to Connecticut residents whose data is included. When privacy notices change to cover new uses like AI training, companies must notify consumers and provide a mechanism to withdraw prior consent.
Connecticut's data breach and safeguards laws require businesses to protect personal information when deploying AI systems and notify individuals of unauthorized access.
Consumer Protection and Unfair Practices
Connecticut's Unfair Trade Practices Act (CUTPA) covers AI-generated advertising that misrepresents product price, quality or characteristics. The state's Department of Consumer Protection and Attorney General enforce CUTPA with broad investigative powers.
Violations carry civil penalties up to $5,000 per violation, plus injunctive relief and restitution. Connecticut consumers can also sue directly for measurable financial losses.
The Connecticut Antitrust Act applies to AI systems used to fix prices, allocate markets or rig bids in coordination with competitors.
Recent Enforcement Actions
The advisory references recent AG enforcement actions against businesses for "misusing algorithms" to addict children and teens and create monopolies in internet search, smartphones and live event ticketing.
Pending Legislation
Connecticut lawmakers are considering four bills that would add AI-specific requirements:
- SB 4 would require data broker registration, restrict personalized algorithmic pricing, define facial recognition technology and prevent sale of geolocation data.
- SB 5 would require AI companion operators to detect and address suicide and self-harm risks, and restrict how automated employment decisions are made.
- HB 5037 would limit notification times for minors on social media and require mental health warnings during sessions.
- SB 435 would add protections for employees subject to automated employment decisions.
All four bills remain in early legislative stages. The Connecticut legislature adjourns May 6, 2026, making passage unlikely before session end.
Your membership also unlocks:
