These answers draw in part from “A Developing Framework for Ethical Use of Artificial Intelligence in Behavior Analytic Practice” by Mahin Para-Cremer, M.Ed., BCBA, LBA (BehaviorLive), and extend it with peer-reviewed research from our library of 27,900+ ABA research articles. Clinical framing, BACB ethics code references, and cross-links below are synthesized by Behaviorist Book Club.
View the original presentation →The Consortium is a professional body that has developed an ethical framework specifically for AI use in ABA practice. Its framework centers on four properties — truthfulness, accountability, transparency, and client welfare — and provides guidance for practitioners and organizations evaluating AI tools. It represents the field's proactive response to AI adoption outpacing regulatory frameworks, and its guidance applies to any BCBA currently using or considering AI-assisted tools in clinical work.
The Consortium represents the kind of proactive professional response that the BACB Ethics Code encourages: rather than waiting for harms to emerge and then developing reactive guidelines, the Consortium is building the framework that allows practitioners to make ethical decisions now, while AI adoption is still in its early stages in the field.
The practitioner who used the tool remains responsible. Under the BACB Ethics Code (2022), responsibility for clinical decisions does not transfer to technology. If an AI documentation tool generates an inaccurate session note and the BCBA submits it without review, the BCBA is accountable for the inaccuracy.
Organizations that deploy AI at scale must establish oversight structures that maintain this accountability chain across every use of the tool. In practice, maintaining the accountability chain requires documentation: records of which AI tool was used, what its outputs were, who reviewed them, what changes were made before use, and what the final clinical product looked like. That documentation is not administrative overhead — it is the mechanism through which accountability is operationalized.
The three primary categories are: documentation AI (which generates session notes or treatment summaries), data analysis AI (which identifies patterns in behavioral data), and communication AI (which generates recommendations or family resources). Each raises distinct ethical concerns. Documentation AI raises accuracy and consent concerns; data analysis AI raises interpretability concerns; communication AI raises truthfulness and competence concerns.
Practitioners should evaluate each category separately rather than applying a single blanket policy. Understanding these categories also helps practitioners prioritize their oversight investments: documentation AI used at scale requires systematic sampling-based review, while data analysis AI used for individual case decisions requires review of every output before it informs treatment. The oversight intensity should match the clinical stakes of each tool's function.
Transparency requires that families understand when and how AI is used in their child's care — not just that a disclosure appears in consent paperwork. Effective transparency involves explaining what the AI does, what data it uses, what errors are possible, and what review process the practitioner applies to AI outputs. Families of individuals with autism and intellectual disabilities have particular reason to expect this level of transparency given historical concerns about consent in ABA contexts.
Transparency should also be treated as an ongoing obligation rather than a one-time disclosure: as AI use evolves, as tools are updated, or as new applications are introduced, families should be informed of those changes in a timely and accessible way. A family's initial consent to AI use does not cover all future uses of AI in their child's care.
Request performance data from the vendor — accuracy statistics, error rates, and information about the populations included in training data. Evaluate whether those statistics reflect real-world conditions similar to your practice. Run a pilot period during which you independently verify a sample of AI outputs against your clinical knowledge.
Pichardo et al. (2026) demonstrated the value of systematic replication in establishing the reliability of clinical data — apply the same standard to AI outputs. Organizations should also run internal accuracy audits periodically — comparing AI outputs against clinically derived conclusions on a sample of cases — and track accuracy metrics over time.
If a tool's accuracy degrades (due to population drift, software updates, or changing clinical contexts), that degradation should be detected early enough to make appropriate adjustments.
AI can assist with documentation if the practitioner reviews and takes responsibility for every output before submission. Time saved on documentation is only ethically neutral if the documentation produced is accurate. AI-generated notes that contain factual errors, mischaracterize client behavior, or reflect assumptions inconsistent with the clinical record create liability and may harm clients.
The time savings do not justify reduced review standards. The oversight requirement also applies to AI tools that are used to generate resource materials, parent guides, or educational content: any material that appears under the practitioner's name or the organization's name must be reviewed for accuracy before distribution, regardless of how efficiently the AI produced it.
AI tools trained primarily on data from white, English-speaking, higher-income populations may produce systematically biased outputs when applied to clients from different backgrounds. This is a direct client welfare concern: biased recommendations or documentation may disadvantage clients who are already underserved. Practitioners should ask vendors explicitly about the demographic composition of their training data and prioritize tools with evidence of equitable performance across the populations they actually serve.
Requesting specific demographic composition data from vendors — and comparing that data to the demographics of the population being served — is a concrete step practitioners can take before adoption. If a vendor cannot or will not provide this information, that is a significant transparency concern that should factor into the adoption decision.
The 2022 Code does not address AI explicitly, but its existing provisions apply fully. The competence requirement applies to AI tool selection and use. The truthfulness and documentation requirements apply to AI-generated content.
The client welfare requirement applies to every decision made about AI adoption. The accountability framework means practitioners cannot disclaim responsibility for harms resulting from AI tools they chose to implement. The Consortium's framework translates these general provisions into AI-specific guidance.
The practical application of existing Code provisions to AI contexts requires interpretation: the competence provision means understanding AI tools well enough to use them responsibly, not necessarily having technical expertise in machine learning. The truthfulness provision means reviewing AI outputs for accuracy before acting on them. The documentation provision means recording AI use in ways that allow the clinical record to be interpreted accurately.
A comprehensive AI use policy should specify: which tools are approved for which functions, who reviews AI outputs and at what frequency, what criteria govern that review, how clients and families are informed about AI use, what data security protocols govern AI-processed information, and how errors or harms attributable to AI are identified, documented, and addressed. Policies should be written before adoption and maintained prospectively, not reconstructed after problems occur. The policy should also specify how clients and families are notified of AI use both at intake and when AI use changes — ensuring that informed consent is a living process rather than a one-time signature on a document that may not accurately describe the current clinical environment.
Follow the Consortium for Ethical Artificial Intelligence in Applied Behavior Analysis for field-specific guidance. Read vendor documentation critically rather than accepting marketing claims. Network with colleagues who are also evaluating AI tools — collective experience produces better evaluation than individual trial-and-error.
Apply single-case research logic (Kok et al. (2026)): treat AI tool adoption as an intervention, measure its effects on clinical outcomes, and make data-based decisions about continuation, modification, or discontinuation. Participating in professional discussions about AI in ABA — through conference presentations, journal articles, and professional organization committees — also contributes to the field-level knowledge base that all practitioners benefit from.
The individual practitioner's experience with a specific AI tool is data that the field can learn from; sharing that experience in appropriate venues supports the collective professional learning that makes ethical AI adoption more likely.
The ABA Clubhouse has 60+ on-demand CEUs including ethics, supervision, and clinical topics like this one. Plus a new live CEU every Wednesday.
Ready to go deeper? This course covers this topic with structured learning objectives and CEU credit.
A Developing Framework for Ethical Use of Artificial Intelligence in Behavior Analytic Practice — Mahin Para-Cremer · 1 BACB Ethics CEUs · $20
Take This Course →We extended these answers with research from our library — dig into the peer-reviewed studies behind the topic, in plain-English summaries written for BCBAs.
279 research articles with practitioner takeaways
258 research articles with practitioner takeaways
252 research articles with practitioner takeaways
You earn CEUs from a dozen different places. Upload any certificate — from here, your employer, conferences, wherever — and always know exactly where you stand. Learning, Ethics, Supervision, all handled.
No credit card required. Cancel anytime.
All behavior-analytic intervention is individualized. The information on this page is for educational purposes and does not constitute clinical advice. Treatment decisions should be informed by the best available published research, individualized assessment, and obtained with the informed consent of the client or their legal guardian. Behavior analysts are responsible for practicing within the boundaries of their competence and adhering to the BACB Ethics Code for Behavior Analysts.