AI Resilience Report forSpeech-Language Pathology Assistants

Right now, AI in speech-language pathology is mostly augmenting the human work, not replacing it — and that's especially good news for Speech-Language Pathology Assistants (SLPAs), who focus on hands-on therapy. The biggest area where AI is taking over routine work is administrative: data collection, progress notes, and documentation. AI-driven tools can help SLPs reduce administrative workload by automating routine tasks.

For instance, speech-to-text systems can quickly convert spoken interactions into written form, streamlining session documentation and minimizing the need for manual note-taking. A large JAMA-published study covered by STAT News ^[1] found those using the technology saved 16 minutes of documentation time and spent 13 fewer minutes in the medical record for every eight hours of patient care — useful, but modest.

For the therapy itself, AI is still far from ready. Stanford researchers testing 15 leading AI models (including GPT-4 and Gemini) on pediatric speech tasks found that the Food and Drug Administration recommends that it be at least 80-85% accurate ^[2]. None of the 15 LMs tested came close to that level of accuracy.

Still, the Stanford team sees AI as a helper: without replacing the important human connection that a clinician provides, an AI system could potentially simplify several of the more tedious steps in this pipeline, freeing SLPs to give children important one-on-one attention ^[2]. The ASHA Leader puts it similarly — these tools should be "accessible, clinically validated, and integrated responsibly—augmenting, not replacing, the human-centered care that is foundational to our profession".

Adoption will likely be steady but cautious. On the "speed up" side, there's a real labor crunch: SLP caseloads are heavy, demand is climbing fast ^[3], and the efficiency brought by these technologies could improve communication access for people with speech and language impairments and help address the rising demand for SLP services, which is fueled by factors such as an aging population, higher survival rates after stroke and cancer, and increased diagnoses of neurodevelopmental and neurodegenerative conditions. Healthcare systems are leaning into ambient scribes because, as Optum Advisory experts note ^[4], more health systems are adopting ambient AI tools to reduce physicians' administrative workloads and give them more time for patient care.

But several factors slow things down. Widespread adoption remains limited due to system-level barriers. Many tools are still in development or not commercially available, and other barriers include disparities in digital literacy, data privacy concerns, algorithmic bias, legal liability, and patient mistrust.

The MDPI Healthcare review ^[5] similarly warns that adoption is challenged by data bias, lack of transparency, and limited integration into clinical workflows. And bias is real — Stanford's testing ^[2] showed the models were better at diagnosing speech issues in boys than in girls, in English speakers than in speakers of other languages, and in older children compared with younger.

The bottom line for young people considering this career: the parts of your job that are most "you" — playing therapy games with a child, encouraging a stroke survivor through articulation drills, picking up on someone's frustration — are exactly the skills AI is worst at. Expect AI to handle more of your paperwork, freeing you to do more of the human work that drew you to the field.