Artificial Intelligence (AI) Considerations: Ensuring Ethical Responsibility

Ensuring ethical responsibility is one key consideration when using generative artificial intelligence (AI) clinically. Make sure you also consider:

It is a clinician’s ethical responsibility to provide competent care and accurate information about professional services to their clients, patients, or students (see Principle III, Rule E of the ASHA Code of Ethics). This includes ensuring that any AI-created content is accurate. Some products use ambient listening and generative AI to create documentation like a treatment note. As a clinician using generative AI, you must know

state and local laws,
payer regulations, and
your facility’s policy or summary of recommendations.

It is very possible that generative AI could produce inaccurate, misleading, content—depending on how you are using that generative AI tool. Before signing or sharing any AI-generated document, be sure to take the following critical steps:

Read and understand the terms of use—including privacy provisions.
Understand the limitations and intended use of the tool(s).
Read through the AI-generated text to ensure that all relevant information is included.
Edit the generated content for errors—add clarifying details (e.g., specific data, personal information) to the finished document.

Examples

See the following use cases of audiologists and SLPs ensuring ethical responsibility when using generative AI tools.

Audiology Example

An audiologist wants to create a template to use for vestibular assessment reports. They create a secure account with a publicly available generative AI tool, input general information about the structure and content that they want included, and prompt the tool to create a customizable template.

Example Prompt: “Hi, <<GenAI>>. You are an audiologist working in an outpatient medical setting. Create a vestibular assessment report. Use text that I can copy and paste into an editable document. Be sure to include space for me to input patient information, assessment data, and recommendations. Before creating this template, please ask me clarifying questions to refine specific details that should be included in the report.”

Notes and Next Steps:

After receiving the AI-generated template, the audiologist reviews it for accuracy and clarity. They ensure that each use of the template is tailored to reflect the specific tests administered and clinical impressions relevant to that individual case.

Speech-Language Pathology Example

An SLP working in private practice wants to create an informational resource for families about reading disorders in children. They use generative AI tools to create text and graphics—with the end goal of sharing vetted information with their clients and families.

“Draft an infographic that I can print or send in an email to share information with families about written language disorders. Consider guidance from www.asha.org to assist with terminology. Use www.asha.org/public/speech/disorders/learning-disabilities/ for additional specific language related to speech services. Use language from https://www.asha.org/practice-portal/clinical-topics/written-language-disorders when describing impacts to functional communication. Use plain language, and consider a list of five FAQs to include in the text. Offer me the text that I can review and edit before offering examples of infographics. Ask clarifying questions so that you can provide a good first draft.”

Notes and Next Steps:

Before entering the prompt, the SLP reviews the tool’s terms of use and privacy policies, specifies trusted sources, and ensures no protected health information is included. After generating the text, they edit and customize it to reflect their clinical expertise, branding, and professional details.

Next: Prioritizing Security