AI Use Across School-Based and Related Service Professionals: What the Data Suggest
Key Points (For Readers on the Go)
A new preprint by my research team examined AI use across 1,350 Ohio school-based and related service professionals.
The data were collected in spring 2025, so the findings are useful but already somewhat dated given how quickly AI adoption is changing.
School psychologists, counselors, and speech-language pathologists reported the highest AI use.
Disclosure and informed consent practices varied across professional roles and were often limited.
Nearly half of respondents reported no AI training, with especially large gaps among assistant-level providers.
The findings point to a clear need for shared AI guidance, training, and governance across multidisciplinary teams.
Introduction
AI use across school-based professionals is no longer a future issue. It is already shaping documentation, communication, report writing, summarization, and professional workflows.
This post is based on our preprint: “AI Use, Disclosure, and Training Across School-Based and Related Service Professionals in Ohio.” The full preprint is available on OSF: https://osf.io/preprints/psyarxiv/w5jpb_v1
One important note: the data were collected between April and June 2025. In AI time, that is not old, it’s ancient. Since then, tools have improved, organizations have started developing more guidance, and many professionals have likely gained more exposure to AI.
Even so, the findings are valuable. They give us one of the clearest early pictures of how AI use differs across school-based and related service professions.
What the Study Examined
The study surveyed 1,350 allied health and related service professionals practicing in Ohio.
Participants included:
Physical therapists
Occupational therapists
Speech-language pathologists
School psychologists
Physical therapist assistants
Occupational therapy assistants
Counselors
Nurses
The study focused on four practical questions:
How often are professionals using AI?
Do professionals disclose AI use to stakeholders?
Do professionals obtain informed consent before using AI?
What type of AI training have professionals received?
These are not abstract questions. They matter for everyday practice.
In schools and related service settings, professionals often share reports, recommendations, intervention plans, progress notes, and summaries. If AI is being used to help create those materials, teams need to understand when, how, and under what safeguards that use is occurring.
AI Use Across School-Based Professionals Is Uneven
One of the clearest findings was that AI use differed by professional role.
School psychologists reported the highest levels of AI use. In the study, 79.1% of school psychologists reported some AI use, and 43.4% reported using AI weekly or daily.
Counselors and speech-language pathologists also reported relatively high use. Counselors reported 77.1% any use, with 28.9% using AI weekly or daily. Speech-language pathologists reported 71.8% any use, with 27.4% using AI weekly or daily.
Use was lower among some other groups. Physical therapist assistants reported the lowest use, with 28.5% reporting any AI use and 7.3% using AI weekly or daily.
This does not mean that some professions are more ethical, innovative, or prepared than others. A more likely explanation is that different roles have different daily work demands.
For example, school psychologists, counselors, and SLPs often produce written reports, family-facing communication, intervention language, and recommendations. These are exactly the kinds of tasks where generative AI tools are most likely to be used.
Assistant-level providers may have less independent responsibility for documentation or may have less access to professional development related to AI.
That distinction matters. AI adoption is not only about individual interest. It is also about role expectations, access, supervision, and organizational culture.
Disclosure Practices Were Limited
Disclosure is one of the most important governance issues in applied AI use.
Disclosure means telling a student, family, client, colleague, or organization that AI was used to support part of the work. This might include drafting a report section, summarizing records, generating recommendations, or preparing educational materials.
In this study, disclosure rates were generally low.
The overall median disclosure rate was 1% of the time. In plain language, many professionals using AI were rarely telling stakeholders that AI had been involved.
There were differences across groups. Counselors reported the highest mean disclosure rate at 41.2%, followed by school psychologists at 29.0%. Nurses and physical therapist assistants reported lower mean disclosure rates, at 13.6% and 14.4%, respectively.
This is not simply a compliance issue. It is a trust issue.
If a family receives an evaluation report, therapy recommendation, or intervention plan, they may reasonably assume that the language was generated entirely by the professional. If AI contributed meaningfully to the product, professionals and organizations need clear expectations about when disclosure is appropriate.
The key question is not, “Should every minor AI use be disclosed every time?”
The better question is: When does AI use become meaningful enough that transparency is ethically important?
That is where many organizations still lack clear guidance.
Informed Consent Remains Unsettled
Informed consent is related to disclosure but not identical.
Disclosure is about transparency. Informed consent is about whether the individual or family agrees to the use after being given understandable information about risks, benefits, and limitations.
Across the sample, 77.4% of AI users reported not obtaining informed consent before using AI in their work.
That finding should get our attention.
At the same time, consent is complicated. Not every use of AI carries the same level of risk.
For example, there is a meaningful difference between:
Asking AI to improve the readability of a general parent newsletter
Asking AI to summarize de-identified professional development notes
Entering identifiable student information into a tool
Using AI to draft eligibility language or recommendations for a psychoeducational report
These are not the same ethical scenario.
The study found small but statistically significant differences across professional roles. Some groups were more likely than expected to report obtaining consent, while others were more likely to report not obtaining consent.
But the larger takeaway is this: many professionals are making individual judgment calls without consistent organizational policy.
That is not sustainable.
Training Gaps Were Clear
The training findings were especially important.
Nearly half of respondents reported no AI training. Only 24% reported formal professional development.
The differences by role were substantial.
School psychologists were more likely to report formal AI training, with 45% reporting formal training. Counselors reported 41%, and speech-language pathologists reported 39%.
By contrast, assistant-level providers reported much lower exposure to training. 76% of occupational therapy assistants and 81% of physical therapist assistants reported no AI training.
Nurses stood out in a different way. They were more likely to report self-directed training, with 38% indicating self-directed learning.
This matters because AI training is not just about learning how to use a tool. It is about understanding:
What information should not be entered into AI tools
How to check AI-generated content for errors
When AI use should be disclosed
When consent may be needed
How bias can show up in AI-generated outputs
How to preserve professional judgment and accountability
Without training, professionals may either avoid AI entirely or use it in ways that create avoidable risk.
Neither outcome is ideal.
Practical Examples for Schools and Related Service Teams
Example 1: A Multidisciplinary Evaluation Team
A school psychologist uses AI to help draft portions of a psychoeducational report. An SLP uses AI to summarize language intervention data. An OT does not use AI at all. The team meets to discuss eligibility and services.
Without shared expectations, team members may not know which parts of the documentation were AI-assisted, whether the outputs were checked, or whether any student information was entered into a tool.
A practical solution would be a team-level AI agreement that addresses:
Which tools are approved
What information may be entered
Which uses require disclosure
Which uses require consent
How AI-generated content must be reviewed
Example 2: A District Professional Development Plan
A district notices that school psychologists and counselors are already using AI, but nurses and assistant-level providers have had little or no training.
Rather than offering one generic AI session, the district creates role-specific examples within a shared ethical framework.
The shared training covers privacy, bias, transparency, and verification. Then each professional group discusses realistic use cases from its own work.
That structure supports consistency without pretending every role uses AI in the same way.
Example 3: Parent-Facing Communication
A professional uses AI to draft a parent-friendly explanation of a service recommendation. The final version is reviewed and revised by the professional before being shared.
This may be a reasonable use case if no identifiable information is entered into an unapproved system and the professional carefully verifies the content.
But the organization still needs guidance on whether this type of use should be disclosed, documented internally, or limited to approved platforms.
The goal is not to create fear. The goal is to create consistency.
What This Means for School Psychologists
School psychologists are in an important position.
They often coordinate multidisciplinary evaluation data, communicate with families, consult with teachers, and help teams make decisions about services and supports.
Because of that role, school psychologists may be well positioned to help their teams ask better questions about AI use.
For example:
Are we using AI in ways that involve student information?
Do we have shared expectations for checking AI-generated text?
Do families need to be informed when AI meaningfully supports a report or recommendation?
Do all team members have access to training?
Are assistant-level providers included in AI guidance and professional learning?
Are our policies keeping up with actual practice?
School psychologists do not need to become AI experts. But they can help ensure that AI use is ethical, transparent, and connected to sound professional practice.
Final Takeaways
AI use is already occurring across school-based and related service professions, but adoption varies by role.
School psychologists, counselors, and SLPs reported the highest use and more formal training.
Disclosure and informed consent practices remain inconsistent.
Training gaps are substantial, especially among assistant-level providers.
Schools and organizations need shared AI guidance that addresses privacy, transparency, human oversight, and role-specific use cases.
Call to Action
If your school, district, clinic, or professional organization is trying to develop practical AI guidance, I can help.
I provide training, consultation, and policy support focused on ethical and responsible AI use in education, psychology, and related service settings.
To learn more, visit Lockwood Educational & Psychological Consulting or contact me through the website.
AI Use Disclosure
Portions of this post were drafted with the assistance of an AI writing tool and revised by the author for accuracy, clarity, and professional judgment.