Citizen Portal
Sign In

Get Full Government Meeting Transcripts, Videos, & Alerts Forever!

Get email alerts on the Artificial Intelligence In Healthcare topic

No spam. Unsubscribe anytime.

Board of Psychology backs limits on AI in therapy, asks for clarifications on AI assessment and workplace rules

California Board of Psychology Legislative and Regulatory Affairs Committee · June 6, 2025
AI-Generated Content: All content on this page was generated by AI to highlight key points from the meeting. For complete details and context, we recommend watching the full video. so we can fix them.

Summary

The California Board of Psychology's legislative committee reaffirmed support for a bill limiting autonomous AI use in psychotherapy and discussed several other AI bills, asking staff for clarifications on assessment uses, informed‑consent language and enforcement responsibilities before taking further positions.

The California Board of Psychology Legislative and Regulatory Affairs Committee on April 24 recommended the full board continue its support for legislation that restricts autonomous artificial‑intelligence tools from delivering therapy without a licensed professional.

Board staff described Senate Bill 9003, a Padilla‑authored measure the committee has already backed, as aimed at protecting consumers by prohibiting unlicensed individuals or entities from delivering therapy via AI, requiring clear patient consent when AI is used as a support tool, and ensuring AI cannot make independent therapeutic decisions. “The amendments clarify enforcement authority and give boards explicit rulemaking power,” staff member Miss Whitney told the committee.

Committee members pressed staff for details about whether certain AI uses—most notably psychological assessments and clinical decision‑support tools—would be covered. “As written today, a psychologist could not make assessments with AI as part of this bill,” said Daniel Phillips, enforcement manager, when asked directly about assessments. Several committee members and commenters urged stronger, explicit guardrails for assessments, saying automated testing poses distinct risks.

The committee also debated AB2575, a separate bill requiring disclosure about AI clinical decision tools and protecting clinicians’ authority to override AI output. Committee members raised concerns the bill’s language currently centers on health‑facility reporting and does not clearly place psychology‑specific duties (such as informed‑consent obligations) on the Board of Psychology. After discussion and requests for regulatory counsel review, the committee voted not to recommend a support position on AB2575 and opted to continue monitoring and to return with additional counsel input.

Public commenters urged a cautious approach. “AI is pretty intrusive,” said Dr. James Willis, noting the technology’s rapid integration into everyday tools. Multiple licensees requested the board convene a broader AI‑landscape review to map risks, enforcement gaps and areas where regulation or guidance is needed.

Next steps: staff will continue to monitor amended language on the bills, prepare more detailed analyses of assessment‑related risks and enforcement pathways, and present updates for full‑board consideration at the May meeting.