ASPPB: AI in Psychology Regulation Across Six U.S. and Canadian Boards

Artificial intelligence (AI) is rapidly becoming a central issue for professionals across disciplines, and its relevance to psychology regulation is particularly consequential. For regulatory boards and colleges, AI introduces complex challenges at the intersection of professional standards, ethical practice, accountability frameworks, and risk management, while also raising new questions about competence, supervision, and the appropriate use of technology in clinical and assessment settings. As AI tools become more embedded in mental health services, regulators are increasingly tasked with ensuring that innovation does not outpace safeguards designed to protect the public.

The Association of State and Provincial Psychology Boards (ASPPB)® gathered six perspectives from regulatory boards and colleges in Arizona, Ontario, Quebec, New Mexico, Minnesota, and Missouri, including individual commentary, legal and ethical analysis, and jurisdiction-specific approaches. Together, they provide insights on how different regulatory bodies and their professionals are responding to the challenges and opportunities posed by artificial intelligence to professional regulation.

Boundaries of Use: AI and Its Legal and Ethical Ramifications in the Regulatory Space

For AI to be effectively integrated into psychological practice, it must be clearly understood, appropriately regulated, and implemented with robust safeguards rather than approached with apprehension. Sarah Ledgerwood, JD, Chief Legal Counsel/Deputy Division Director of the Division of Professional Registration, emphasizes the importance of developing a deeper understanding of AI’s role and implications for professional regulation.

Drawing on 17 years of regulatory experience, Ledgerwood has observed significant shifts in the landscape, from the rapid expansion of virtual practice during the COVID-19 pandemic to the emerging challenges posed by artificial intelligence. She notes that the acceleration of virtual processes during the pandemic offers a useful parallel to the current moment, as boards and colleges once again confront the challenge of adapting regulatory frameworks in response to transformative technological change.

“We have had a lot of changes in in-person versus virtual supervision and even care. So I think it’s in the same vein. As regulators, we have to accept AI is here and we have to incorporate it into our ethics and what we expect of our practitioners,” said Ledgerwood. Understanding the legal and moral implications of AI is fundamental for her and her colleagues from other jurisdictions.

According to a written statement from the Ordre des Psychologues du Québec, “regulators in Québec have no prohibition on using AI to support professional practice,” and the Order takes a non-prescriptive approach and promotes the use of clinical judgment and professional autonomy. For them, “ethical and legal obligations—including confidentiality and the protection of personal information—remain fully applicable regardless of the tool used.” The Order promotes responsible use of AI as a clinical tool, not as a substitute.

In contrast to Québec’s position, the Arizona Board of Psychologists Examiners had not endorsed the unrestricted use of AI in psychological practice. However, both boards see eye to eye on viewing AI as an assistive tool. Arizona, in particular, has not endorsed the use of AI in psychological practice for diagnosis, risk assessment, administration of psychological tests, or other functions that require nuanced clinical judgment. Current Arizona Board discussions generally emphasize viewing AI as an assistive tool rather than a substitute for clinical expertise.

Equally important to psychology regulators are matters such as professional accountability, boundaries of use, validation standards, documentation requirements, and AI literacy expectations. However, regulators agree that responsibility for the responsible use of AI lies with practitioners.

This is where the concept of augmented intelligence comes into play, as it refers to the use of AI tools designed to assist—not replace—human clinicians. By focusing on enhancing human reasoning, pattern recognition, and empathy, augmented intelligence could reduce burnout and administrative burdens while allowing licensed psychologists to retain control over final patient decisions.

Responsibilities of Psychology Regulators: Accountability

Tony DeBono, MBA, Ph.D., C.Psych., and Registrar & CEO of the College of Psychologists and Behaviour Analysts of Ontario, is keen on human accountability and on balancing the potential benefits and risks of AI use by the College and for psychologists.

“The role of technology and its idea of augmented intelligence that a practitioner will have is a consideration. But ultimately, it is incumbent upon that individual to use their knowledge, skills, and judgment to safely use whatever technology is available,” said DeBono. “We cannot defer responsibility to the technology to say, ‘Oh, well, it was AI that did it.’ Ultimately, it’s the human beings behind the technology, behind the machines, that have to account for it.”

The duty of Boards and Colleges is to regulate the psychology profession, not the companies that develop AI. This is fundamental, as all six participants from Arizona, Ontario, Quebec, New Mexico, Minnesota, and Missouri agree that the presence of AI in clinical practice does not shift accountability from the licensed psychologist.

Meranda Marrin, PhD, Board Member of the New Mexico State Board of Psychologist Examiners, stated that the Board focuses on whether the psychologist exercised independent professional judgment, as this aligns with the Board’s rules and statutes. “We are going to want to know if they really understood the tool’s limitations, if they appropriately reviewed the output, whatever the output is. And then, of course, are they operating within the bounds of competence?”

Diana Medina, PhD, Board Chair for the Arizona Board of Psychologists Examiners, offered her personal professional perspective on how regulators could determine decision-making in cases where AI played a role. “From a regulatory perspective, the presence of AI in clinical practice does not shift accountability away from the licensed psychologist.” Ongoing Board discussions have generally emphasized that AI should be viewed as an assistive tool rather than a substitute for clinical expertise.

The Ordre des Psychologues du Québec enlisted some factors to be considered when assessing the psychologist’s responsibility, including:

1. The psychologist’s competence and training in using the tool,
2. Whether informed consent was obtained,
3. The validity and reliability of the tool,
4. Human verification of AI-generated output,
5. Consideration of the tool’s biases and limitations, and
6. The potential for harm to the client.  

Using AI tools is inherently linked to third-party providers, raising concerns about documentation and responsible data management.

AI and Documentation Requirements

Patients’ Rights are fundamental, particularly to expect that information provided to the psychologist is kept in the strictest confidence. In Minnesota, there is no current requirement to disclose AI in documentation, at least not from any legislative authority. “In terms of transparency in the psychologist-client relationship, up front informed consent, Minn. R. Ch 7200.4720 Subp. 1F requires informed consent on information and uncertainty of benefits if the proposed service or method is innovative in nature.” However, as per Sands and Markanda, the Board has not yet determined what, if any, type of disclosure is necessary in various forms of clinical documentation, such as the authoring of a particular progress note.” 

The New Mexico State Board of Psychologist Examiners has not specified how to include AI in documentation; however, as Dr. Marin pointed out, their rules require that records include diagnoses, clinical formulations, test evaluation services or results, any consultations, and evaluation reports. “I think, within that, what we would be looking at is whether there is documentation to show what role AI played and whether the psychologist conducted an independent review.”

The Ordre des Psychologues du Québecpointed out that “free and informed consent is required whenever personal data is provided to an AI tool.” They list factors that should be considered when obtaining patient consent:

1. The type of AI tool used,
2. The purpose for which it will be used,
3. The data retention period,
4. The confidentiality protocol and who can access the information,
5. The limitations and risks associated with the AI tool.

For the Ordre, it is important that, for the sake of transparency and to maintain trust, the psychologist should indicate in their clinical records whether AI was used and for what purposes.

Just as important as disclosing the use of AI to clients is the psychologist’s responsibility to carefully verify the output generated by AI-assisted note-taking tools, given the risk of AI hallucinations. These occur when an AI system confidently produces content that is incorrect, altered, omits facts, or entirely fabricates data, rather than accurately reflecting the original notes or transcript.

For Ledgerwood, it is important to begin educating graduate students about AI and its risks and opportunities, particularly because AI is inherently part of their world, with accountability for using AI apps being fundamental. “AI is inherently wanting to please you, and I have never even thought about it because I have used it in a very limited way.” However, that is not the case for future licensed psychologists. “I think we need to do some work with talking to our graduate students, because this is going to be their life.”

Furthering public protection is fundamental to all boards and colleges alike, creating a significant opportunity for regulators in the era of AI: Continuing Education.

Literacy Expectations for Psychology Regulators

Continuing education is important for psychology regulators as they continue to work towards public protection, and they also need to deepen their understanding of AI. “I think we are in the same position as the profession itself: learning rapidly while attempting to keep pace with technological change,” said Dr. Medina. “Many boards have begun engaging in ongoing discussions, conference presentations, legal consultations, ethics reviews, and interjurisdictional collaboration regarding AI. Organizations such as the Association of State and Provincial Psychology Boards (ASPPB) have been particularly valuable in helping regulators, especially me, share information and identify best practices across jurisdictions.”

For Dr. Medina, regulators do not necessarily need to become technology experts, but they do need enough understanding of evolving AI tools; “AI literacy should become part of the ongoing professional competence of psychologists themselves.”

With the New Mexico Board of Examiners, Dr. Marin points out that there are opportunities for psychologists to pursue continuing education credits focused on the use of AI; these credits could cover ethics, privacy and confidentiality, data storage, and protection.

In addition, the Ordre des Psychologues du Québecpoints out that its regulators are reinforcing their knowledge of AI through training sessions and discussions with members, the Ordre president’s participation in panels and discussions on the use of AI in clinical practice, and by informing its members in the Ordre’s magazine.

The presence of AI and its impact on regulatory boards and colleges is summarized by Dr. DeBono: “AI has significant potential to increase efficiencies and expand the knowledge base by processing large volumes of psychological data to identify trends.” However, DeBono’s optimism is tempered by the pace of AI advancement, which means that possibilities today may be far exceeded in the coming months and years.

These are important considerations for governing bodies that must remain flexible and adjust parameters as AI continues to evolve in sophistication and they steadfastly advance in their duty to protect the public.