AI Is Changing and So Are the Rules
Adapting to the changing AI landscape

AI Rules Are Becoming More Practical

Beginning January 1, 2027, Colorado's updated artificial intelligence law (S.B. 26-189) will take effect. The new law significantly narrows and refines the state's original generative AI legislation, shifting the focus from broad regulation to responsible AI use, particularly in decision-making technology (Colorado Attorney General, n.d., S.B. 26-189, 75th Gen. Assemb., Reg. Sess. 2026).

What This Means for Providers

For specialists such as speech-language pathologists and psychologists, the message is straightforward: Generative AI can be a valuable tool for improving efficiency, but it should never replace professional judgment. AI may help draft, edit, or rewrite documentation, but providers remain responsible for the decisions and rationale presented in every report, conclusion, recommendation, and intervention strategy.

Transparency is equally important. Families, caregivers, and providers should understand how AI will (and will not) be used during the clinical process, and consent may be required (ASHA, n.d.).

It's important not to confuse preprogrammed tools with generative artificial intelligence. Many assessment platforms, such as Pearson's Q-global, support online test administration by monitoring basal and ceiling rules, calculating scores, and, with some evaluations, providing an automated narrative of results (Pearson Assessments, n.d.). These systems assist the clinician but do not interpret findings or make professional decisions. This distinction reminds us that technology has long been part of the evaluation process. While AI may be the newest tool in the clinical toolbox, it requires more consideration before use.

Start with the Task, Not the Technology

Responsible and ethical generative AI use also means recognizing when it may not be the best tool for the task. Consistent with the new Colorado law, generative AI can be valuable for generating, summarizing, or drafting content, but because it creates new output, that content requires full professional review and validation.

However, for information written and pre-vetted by a provider, a more practical choice may be predictable technology, such as deterministic automated platforms, that can deliver expert-written information consistently without generating new verbiage.

Of course, when a task requires interpretation, clinical reasoning, or professional decision-making, that responsibility should remain with the provider. The goal is not to use generative AI whenever possible; rather, it is to choose the technology that best supports the task while keeping professional expertise at the center.

What Providers Can Do Today

As providers navigate a rapidly changing technology landscape, understanding which tools they are using (and what those tools actually do) is becoming increasingly important. Key steps that providers can take include:

• Know when generative AI is part of the tool or feature
• Understand how and where data is being processed (and be able to explain that to clients and caregivers)
• Review and validate all AI-generated content
• Never rely on AI to replace clinical judgment
• Become familiar with the laws and professional guidelines that apply to practice
• Always follow security protocols consistent with ethical responsibility (ASHA, n.d.)

Colorado is among the first states to establish specific requirements, but it certainly won't be the last. Ultimately, professional responsibility remains with the provider. Before signing any document containing AI-generated content, ask yourself: "Would I be comfortable putting my name and professional license behind every statement?"

Citations and Further Information

• American Speech-Language-Hearing Association. (n.d.). Artificial intelligence (AI): Considerations for CSD professionals. Retrieved August 15, 2026, from ASHA.
• American Speech-Language-Hearing Association. (2023). Code of ethics. ASHA Code of Ethics.
• Autio, C., Schwartz, R., Dunietz, J., Jain, S., Stanley, M., Tabassi, E., Hall, P., & Roberts, K. (2024). Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST AI 600-1). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.AI.600-1.
• Colorado Attorney General. (n.d.). Colorado anti-discrimination in AI law (ADAI) rulemaking. Retrieved August 15, 2026, from Colorado Attorney General.
• Darch, J. E., Brooks, P., Daudi, S., Hinckley, D., & Kaye, A. (2026, May 19). Colorado scales back AI law, with targeted implications for health care. Ropes & Gray. Ropes & Gray.
• Pearson Assessments. (n.d.). CELF-5: Clinical Evaluation of Language Fundamentals, fifth edition. Retrieved August 15, 2026, from Pearson Assessments.
• Pearson Assessments. (n.d.). WISC-V: Wechsler Intelligence Scale for Children, fifth edition. Retrieved August 15, 2026, from Pearson Assessments.
• S.B. 26-189, 75th Gen. Assemb., Reg. Sess. (Colo. 2026). Colorado General Assembly.

About the Authors