Artificial intelligence is no longer on the horizon. It’s here. And for those of us in applied behavior analysis (ABA), it’s arriving embedded in everything from practice management platforms to documentation tools and electronic health record systems. But there’s an urgent issue that remains mostly unspoken: AI systems can violate the most basic human rights of the people we serve—if we let them.

This isn’t just a technological conversation. It’s an ethical one.

Why Privacy Isn’t Just a Legal Checkbox—It’s Foundational

Privacy is not a luxury. It is the bedrock of liberal democratic society. In the United States, we take for granted that individuals have a right to keep certain aspects of their lives out of public view, especially when those aspects involve sensitive information like health, behavior, or an individual’s developmental history, such as assessments, milestones, and intervention outcomes related to learning, communication, or independence.

That expectation is not arbitrary. It’s woven into HIPAA, FERPA, and the BACB’s Ethics Code for Behavior Analysts (2020), which demand that we protect client confidentiality, obtain informed consent, and act in the best interest of those we serve. These are not vague ideals. They are codified obligations.

But today’s technology landscape has shifted the terrain. We’re not just talking about locked file cabinets or encrypted servers. We’re talking about systems that can ingest thousands of data points, generate synthetic outputs, and do so at scale with or without our awareness of how they work. When combined with the ease with which anyone can copy and paste PHI into a browser window to ease effort and receive a half-baked AI-generated output, the current privacy landscape is a tinderbox waiting to explode.