APA Guidance on Evaluating AI Tools for Psychological Practice: A Quick Dive

Want to stay informed about the rapidly evolving intersection of AI and school psychology practice? Sign up for my weekly newsletter where I explore these emerging issues and their implications for our field.

Artificial intelligence (AI) tools are becoming more prominent in the practice of psychology. As these tools become more integrated into our daily professional tasks, it's crucial to know which tools are right for your needs.

The American Psychological Association (APA) has put together a valuable resource titled 'Steps to evaluate an AI-enabled clinical or administrative tool.' This guide is the best I have found for helping mental health professionals evaluate AI tools effectively and make informed decisions about integrating technology into their practice. Below are my main takeaways regarding what any company providing AI technology for psychologists should offer to help you make informed decisions about integrating such tools into your practice. When considering using an AI-enabled psychology tool a practitioner or organization should consider the following:

Company Leadership: Does the company have a psychologist or mental health professional on the leadership team? This ensures that the tool is developed with clinical expertise and aligns with the needs of psychological practice.

Tool Functionality: Does the tool integrate with your current software? Does it streamline your workflow and save administrative time? Is it cost-effective? Does the company provide a demo or a free trial? What kind of technical support is offered (e.g., 24/7 support)?

Clinical Evidence: Is there empirical evidence to support the effectiveness and safety? For example, AI tools can be FDA-cleared digital therapeutics. If not, has the company conducted its own research that you can review? Look for independent research to ensure objectivity.

Legal Compliance: Does the AI comply with HIPAA, and/or FERPA? Additionally, do they offer a business associate agreement (BAA)? Note that some states (e.g., California) may have their own specific laws.

Data Security: Does the company have a clear data security policy? Are the data encrypted? Does the company have other certifications (e.g., HITRUST, SOC 2)? Is the policy clear and readily available?

Privacy Policy: Is the privacy policy available and easy to understand? Read the privacy policy in full. What data are collected? With whom can the data be shared? How long are data retained? Data are often used for model training, employee education, research, and marketing. Can you “opt-out” of some types of data sharing (e.g., for marketing purposes)? While selling de-identified data is not a violation of HIPAA, would your stakeholders provide consent for this? Consider whether this aligns with your practice's privacy standards. Again, some states (e.g., California, Nevada) have specific data privacy laws. Do you have the right to request, delete, or correct data?

Terms of Service (TOS): Is the TOS available for review before using the product? Read the TOS in full. This section should discuss how personal identifiable information (PII) is stored and maintained.

Informed Consent: Does the company provide a sample informed consent document and require an attestation that informed consent has been obtained prior to using the tech with PII?

Contact the Company if You Have Questions: If there are any unanswered questions after your review, reach out directly to the company. They should be willing to provide the necessary information to address your concerns. Consulting with a local attorney can also help clarify complex policy provisions.

Document Your Review: It's essential to document your initial evaluation of the tool. This demonstrates that you have conducted due diligence when selecting an AI-enabled tool. The APA provides an awesome checklist that you can use for this purpose. 

Review Policies for Updates: Policies and TOS are likely to change over time. Regularly reviewing these documents ensures that you remain compliant with any updates or changes. 

Evaluating AI tools doesn’t have to be overwhelming—thanks to the APA's guidance, you can ask the right questions and ensure the tools you use are aligned with best practices for security, privacy, and ethical care. This guide is an invaluable starting point for evaluating AI-enabled clinical and administrative tools.

Stay connected with the latest developments in AI and school psychology! Sign up for my weekly newsletter where I dive deep into these important issues, provide practical guidance, and explore how we can thoughtfully integrate AI while maintaining the highest standards of practice in school psychology. Join our community of forward-thinking school psychology professionals as we navigate this transformative period in our field.

Next
Next

AI in School Psychology Training: Insights from NASP's Latest Brief