The behavior and analyst advisory committee of the Kansas Behavioral Sciences Regulatory Board on Dec. 12 spent substantial time reviewing licensee survey results about artificial intelligence and heard national practice‑parameter guidance from CAASPP, the national association speaker said.

The committee learned that the board’s draft survey found most responding licensed behavior analysts reported not using AI in clinical practice, but a meaningful minority reported AI use for tasks such as email drafting, scheduling, brainstorming and session documentation. David Feige, executive director of the BSRB, said 68 licensed behavior analysts answered the LBA question and 49 reported they do not use AI; 13 reported some AI use. Feige pointed committee members to the full responses in the report appendices.

Summer, a CAASPP representative, told the committee many users may not realize they are using AI when a platform offers suggestive edits or automated summaries. "People may be using AI and not realize that they're using AI," she said, and recommended that organizations adopt clear policies that identify permitted uses and specify who is responsible for verifying AI‑generated material.

Why it matters: Committee members raised privacy and safety concerns tied to AI use — including whether AI tools retain sensitive information or retrain models using provider data, and how AI errors or 'hallucinations' in documentation could affect billing or clinical decisions. CAASPP urged that providers ask vendors explicit questions about data storage, opt‑out options and whether the vendor's product marks AI‑generated content so clinicians can verify accuracy.

What was proposed: CAASPP advised the board to consider vendor‑question templates and practice parameters (an annotated glossary accompanies CAASPP’s guidance) so Kansas licensees have concrete examples and language for organizational policies. Committee members also suggested BSRB staff gather examples of other states' rules; CAASPP offered to introduce the committee to contacts who track licensure practices nationally.

On verification and auditing, CAASPP recommended a human‑in‑the‑loop approach for documentation: while a 100% clinician review of AI‑generated notes is the gold standard, auditing cadences may be tapered as a vendor or tool demonstrates reliability. Summer said organizations should decide audit frequency according to risk and capacity.

Next steps: BSRB staff will include CAASPP’s practice parameters and the packet of state examples in materials for the February meeting, and committee members recommended the advisory committee continue AI regulation discussions in 2026 with possible recommendations to the full board.