HB 4770 West Virginia House of Delegates · 2026 Regular Session

Establishing limitations on the use of artificial intelligence and artificial intelligence technology to deliver mental health care, with exceptions for administrative support functions

HB 4770 prohibits the use of AI companions (systems designed to simulate human relationships through personalized, ongoing emotional interactions) in delivering mental health therapy services. It requires explicit, written consent for any AI use in mental health care, banning such tools from diagnosing, treating, or claiming to treat conditions. The law applies to all health insurance policies issued or renewed after January 1, 2027, affecting mental health providers and insurers. Exceptions allow AI for administrative tasks (e.g., scheduling) but not for direct therapy or emotional engagement.
Bill status in committee 1 of 4 stages cleared
Introduction
Jan 2026
Committee Review
Floor Vote
Governor
Introduced Jan 23, 2026 Last action Feb 3, 2026
Maddy AI version diff · 1 comparison

What changed between versions

Introduced Version Committee Substitute · 7 edits
MODERATE
The bill was amended to remove the creation of a large, multi-agency Task Force on Artificial Intelligence, replacing it with a direct focus on regulating AI in mental health care. The scope was narrowed from general AI policy across the state to specific limitations on how AI can be used to deliver therapy and psychotherapy services. New definitions were added to clarify terms like 'AI companion' and 'therapeutic communication,' and strict rules were established requiring human oversight for all clinical decisions and explicit consent for recording sessions.
Scope change
The bill's scope was significantly narrowed. The original version created a broad Task Force to study AI across all state agencies and industries. The substitute version removes this Task Force entirely and focuses exclusively on regulating the use of AI in the delivery of mental health care.
REQUIREMENT

The creation of the West Virginia Task Force on Artificial Intelligence and its extensive membership list were removed.

The requirement for the Task Force to make recommendations on general AI policy, workforce impact, and public interest use cases was removed.

New rules require operators to notify users they are not communicating with a human during AI interactions and mandate explicit written consent before using AI to record or transcribe therapy sessions.

Strict prohibitions were added preventing AI from making independent therapeutic decisions, directly interacting with clients for therapy, or generating treatment plans without human review.

The amendment clarifies that peer support, religious counseling, and digital mental wellness services cannot use AI to diagnose conditions or create treatment plans without a licensed professional's approval.

DEFINITION

New definitions were added for 'AI companion,' 'AI model,' 'consent,' 'digital mental wellness service,' 'emotional recognition algorithms,' 'generative artificial intelligence,' 'licensed professional,' 'peer support,' 'religious counseling,' 'self-harm,' 'therapeutic communication,' and 'therapy or psychotherapy services.'

ENFORCEMENT

A civil penalty of up to $10,000 per violation was added for operators or licensed professionals found in violation of the new AI restrictions.

Floor votes

How they voted

No floor votes recorded yet.
Full legislative history

Actions timeline

Total actions
7
Key actions
0
Committee
4
Feb 3, 2026
Committee
To House Finance
lower
Feb 3, 2026
Committee
By substitute, do pass, but first to Finance
lower
Jan 23, 2026
Committee
To House Health and Human Resources
lower
Jan 23, 2026
Introduced
Introduced in House
lower
Jan 23, 2026
Committee
To Health and Human Resources then Finance
lower
1 primary · 2 co-sponsors

Sponsors