Integrating AI into Psychology

As technology advances, artificial intelligence (AI) is becoming a valuable tool in various fields, including psychology. AI-assisted tools can enhance therapy, provide new insights, and streamline administrative tasks. However, integrating AI into psychology practice requires careful consideration of ethical and legal guidelines to protect clients' privacy and confidentiality. The Australian Psychological Society (APS) Code of Ethics provides one framework to scaffold psychologists’ decision-making in their adoption of AI. 

Confidentiality, Consent, and AI: What You Need to Know

It is assumed that psychologists will input personal and sensitive healthcare information into AI tools, so understanding when and how this information can be disclosed is crucial. According to section A.5. of the APS Code of Ethics, psychologists can only disclose confidential information with the Client's Consent.

Before using AI tools, clients must give their explicit and informed permission for their data to be processed by these tools. 

Conversations about informed consent should cover topics including, but not limited to:

Purpose and Benefits:

• Explain the primary purpose of using AI tools in psychological services (e.g., to improve accuracy in note-taking and more personalised interventions). 
• Describe the potential benefits of AI tools for their treatment.
• Explain any secondary purpose of using AI tools, which refers to using client data for reasons other than the original intent, often for broader applications that may not directly benefit the individual client (e.g., de-identified data being used to trait AI models or improve service delivery). 

Types of AI Tools:

• Detail the specific AI tools that will be used and how these were selected.
• Explain how these tools will assist in therapy or assessments.

Data Collection:

• Outline what data will be collected by the AI tools.
• Specify how the data will be collected (e.g., therapy session transcripts, letters, assessments, online forms and questionnaires etc).

Data Usage and Storage:

• Describe how the collected data will be used and won’t be used.
• Explain any analysis that AI will perform on the data.
• Explain how long and in what formats the data will be stored.

Confidentiality and Privacy:

• Clarify how confidentiality will be prioritised.
• Explain any measures in place to protect client privacy.

Data Sharing:

• State if and when data might be shared with other professionals or services.
• Explain the circumstances under which data might be disclosed (e.g., legal requirements, risk of harm, third party applications).

Consent and Authorization:

• Obtain explicit consent for the use of AI tools.
• Ensure the client understands they can withdraw consent at any time.

Foreseeable Risks and Limitations:

• Discuss any potential risks or limitations of using AI tools.
• Explain what steps are taken to mitigate these risks.

Client Rights:

• Inform clients of their rights regarding their data.
• Provide information on how they can access, correct, or request deletion of their data.

Transparency and Accountability:

• Ensure clients know how to raise concerns or ask questions about AI use.
• Describe the procedures for addressing any issues related to AI tools.

Legal and Ethical Considerations:

• Highlight the relevant ethical guidelines and legal requirements being followed.
• Ensure clients understand the ethical commitment to their wellbeing and data protection.

By adhering to the APS Code of Ethics, psychologists can leverage AI to enhance their practice while safeguarding the trust and confidentiality that are foundational to the therapeutic relationship.  Understanding these ethical considerations helps both psychologists and clients feel more confident about the integration of AI into mental health care.