Home Stories About Search RSS Feed
Policy 2 min read

Connecticut Passes Sweeping 71-Page AI Regulation Bill Covering Chatbots, Employment, and Frontier Models

Back to News

Connecticut has become the latest — and arguably most ambitious — state to pass comprehensive AI legislation. The 71-page bill, which cleared the state Senate this week, establishes a sweeping regulatory framework covering everything from consumer-facing chatbots to frontier model development.

What the Bill Covers

The legislation is notable for its breadth. Key provisions include:

Companion Chatbot Regulation

  • Mandatory disclosures that users are interacting with an AI system
  • Restrictions on emotional manipulation techniques designed to create dependency
  • Special protections for minors, including parental notification requirements and limits on conversational patterns that could impact mental health

AI in Employment

  • Transparency requirements for employers using AI in hiring, promotion, and termination decisions
  • Right to human review for any adverse employment decision made or substantially influenced by AI systems
  • Mandatory bias audits for AI tools used in employment contexts, with results made available to affected individuals

Frontier Model Developer Obligations

  • Pre-deployment risk assessments for models above a defined capability threshold
  • Incident reporting requirements for safety-relevant failures or unexpected behaviors
  • Record-keeping obligations for training data sourcing, red-teaming results, and capability evaluations
  • Whistleblower protections for employees at AI labs who report safety concerns

The National Context

Connecticut’s bill arrives amid a wave of state-level AI regulation across the US:

  • California and Colorado are advancing bills on healthcare AI and employment automation
  • Florida is convening a special session on April 28 to debate an “AI Bill of Rights”
  • The federal government has yet to pass comprehensive AI legislation, leaving states to fill the vacuum

Industry Reaction

The response has been predictably mixed. AI companies and trade groups warn that a patchwork of state regulations creates compliance complexity and could slow innovation. Consumer advocates and labor organizations argue that federal inaction makes state-level regulation not just appropriate but necessary.

What makes Connecticut’s bill significant is its specificity. Unlike broader framework bills in other states, this legislation includes concrete requirements with defined penalties — making it one of the few state AI laws with real enforcement teeth.


Source: jdsupra.com, ctmirror.org

Evelyn Vance
Written By

Evelyn Vance

Senior Policy & Business Editor

Evelyn Vance has covered technology policy, artificial intelligence regulation, and corporate governance for over a decade. Her work focuses on the intersection of government policy, antitrust, and frontier AI corporate developments.