Ethical Behavior Analysis in the Age of Artificial Intelligence (AI): The Importance of Understanding Model Building while Formal AI Literacy Curricula are Developed
Start slipping basic model-building demos into supervision today—ethical AI practice depends on knowing how the code thinks.
01Research in Context
What this study did
Cox (2025) wrote a position paper. It tells behavior analysts to learn how AI models are built before using the tools.
The paper says formal AI classes do not yet exist in most ABA programs. Therefore, supervisors and professors must add lessons now.
What they found
The paper finds that ethical AI use is impossible without basic model literacy. If you do not know how the math works, you can not spot bias or harm.
No numbers are given. The claim is logical, not statistical.
How this fits with other research
Oleribe (2025) extends the same idea. That paper says GenAI can cut neuro-developmental diagnostic delay by ninety percent. Cox warns you must first understand the model; Oleribe shows the reward once you do.
Root et al. (2020) used a similar teaching move. They wed Skinner’s programmed instruction to online college modules. Both papers push tech-enhanced coursework, just swapping AI for older teaching machines.
Britwum et al. (2020) faced the same hurry. They moved parent training to telehealth overnight and still kept ethical safeguards. Cox asks for the same fast pivot: teach AI literacy before the official syllabus arrives.
Why it matters
You will use AI for data plots, report writing, or client apps soon. If you cannot explain how the algorithm reaches its answer, you risk harming clients and violating our ethics code. Add one short demo to your next supervision meeting. Show how a simple decision tree splits data. Five minutes now prevents big mistakes later.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Open a free AI sandbox, load a toy data set, and walk your supervisee through one split node.
02At a glance
03Original abstract
Ethics is fundamentally concerned with claims of “right,” “wrong,” “good,” “bad,” and how we might know those claims are accurate. Artificial intelligence (AI) is a term that represents a suite of tools where nonbiological systems process data and information to generate an output considered “intelligent.” As with any suite of technologies, choosing the right tool requires the tool user to critically evaluate which tool is best for the task (if one exists). Harnessing the power of AI systems to maximize benefit and minimize harm requires basic AI literacy. AI literacy requires a basic understanding of how mathematical models function. All future behavior analysts will need to be AI literate. This will require changes to education and training programs to ensure students have a basic understanding of model building, especially as we wait for the scholarship and research to unfold that outlines AI literacy skills specific to behavior analysts.
Perspectives on Behavior Science, 2025 · doi:10.1007/s40614-025-00459-z