Applications of Artificial Intelligence (AI) in School Psychology

Written in late 2023, this paper looks at where artificial intelligence could genuinely help school psychology and where it should not be trusted — and spends about as much space on the second question as the first.

The promise it identifies is efficiency and personalization. Turning interview notes into report narrative removes what the field generally agrees is the most tedious part of the work. Synthesizing academic, behavioral and historical data could support more specific recommendations than an IEP team typically produces, and the research it cites finds targeted recommendations more useful than generic ones. Current applications reviewed include risk screening from existing academic data, AI-assisted therapy models where a chatbot supports rather than replaces a clinician, data organization, and progress monitoring across student, school and district levels.

The longer-term section is more speculative — automated documentation, AI-proposed assessment questions, adaptive testing that administers only items at the edge of a student's competence — and the paper is candid that diagnostic accuracy in mental health tops out around 80% because of comorbidity and overlapping presentations.

Then it turns to limits, which is where its value sits: student data privacy and cyberattack exposure, informed consent that families can actually understand, bias inherited from training data, the difference between correlation and the causal explanation a psychologist is after, and the accountability gap — the psychologist remains responsible, the system does not. Its position is human-in-the-loop: AI proposes, a professional decides.