Using AI in counseling practice

“Use artificial intelligence (AI) within your Counseling practice” (NBCC, 2027 task 6.19)

Artificial intelligence now appears in counseling practices as note takers, chat assistants, scheduling tools, screening aids and client-facing apps. Using AI responsibly means protecting privacy, obtaining consent, checking accuracy and keeping clinical judgment with the counselor. The exam tests the ethical and legal limits of AI use.

Key points

Use AI tools only for purposes you understand and can supervise; the counselor remains responsible for assessment, diagnosis, treatment and documentation.
Never enter identifiable client information into AI tools that lack HIPAA safeguards and a business associate agreement.
Obtain informed consent before using AI that records, transcribes or analyzes sessions, and offer an alternative if the client declines.
Review AI-generated notes and summaries for accuracy, bias and omissions before signing them.
Do not rely on AI for risk assessment or diagnosis; it may support but never replace clinical judgment.
Be transparent with clients about any AI they interact with, such as chatbots used between sessions, and its limits, especially in crises.
Laws, payer policies and professional guidance on AI are changing, and some states restrict AI use in mental health services; stay current.

In practice

A counselor uses an AI note tool covered by a business associate agreement. After each session, she reads the draft, corrects a misheard medication name, removes an inaccurate statement about the client's mood, and signs the note.

A client asks if he can use an AI chatbot for support between sessions. The counselor discusses its limits, including that it is not a crisis service, reviews his safety plan and crisis numbers, and asks him to bring anything concerning to session.

Check yourself

Who is responsible for the accuracy of an AI-drafted progress note?

The counselor who signs it, so it must be reviewed and corrected before signing.

Can a counselor enter client details into a free public AI chatbot to get treatment ideas?

No, not with identifiable information, because the tool lacks HIPAA safeguards and the disclosure breaches confidentiality.

Should AI be used to make suicide risk determinations?

No, AI may support but must not replace clinical risk assessment and judgment.

NBCC's knowledge and skills for Domain 6

Knowledge of: State laws (state licensure law, supervision requirements, abuse reporting, license renewal, CAQH); Federal laws (e.g., HIPAA, FERPA, immigration policy, HI-TECH, ADA, NPI); Codes of ethics; Third-party reimbursement, insurance and billing (e.g., CPT codes); Counseling Compact; Emerging technologies in Counseling; Crisis policies and procedures; Recording keeping and reporting requirements and procedures; Appropriate logistics for telemental health services; Scope of practice (e.g., letters of support, working with special populations); Alternative medical-assisted/pharmacological interventions; Accessing state licensure board; Executing voluntary and involuntary commitments.

Skill in: Maintaining confidentiality; Discerning reputable training providers and programs; Troubleshooting technological difficulties during telemental health session (e.g., session disconnects, video not working); Billing and third-party reimbursement.

A study guide written by ExamDragon from NBCC's published task; not NBCC material and not part of the exam.