CoFABA Conference - 2026
Igniting Momentum: Innovation
4/26/20261 min read
Reflections from CoFABA 2026: The Intersection of AI and Behavior Analysis
Last week, I attended the CoFABA 2026 conference in Tampa. As someone deeply embedded in both Software Engineering and Data Science research, it was a privilege to see how our field is navigating the "Digital Age" of clinical practice.
Here is a summary of my key learning tracks from the event:
1. The AI Frontier: "Shadow AI" vs. Clinical Validation
One of the most striking sessions was by Dr. David Cox on the ethics of AI. We are at a crossroads where "Shadow AI" usage is already >80%, yet we lack peer-reviewed validation studies for clinical documentation in ABA-specific tools.
The Risk: Without oversight, LLMs can perpetuate biases (e.g., the "bank robber" bias in autism-related prompts).
The Goal: Moving toward an LLM-driven longitudinal clinical co-pilot that assists practitioners without losing the human element.
2. Advanced Tech Stacks in Behavioral Research
I spent time exploring the technical implementation of "Digital Twins" (like the Living Mirror adolescent twin) and how we can use modern tools to track clinical progress.
The Stack: Integrating AssemblyAI, Receptiviti, Cursor, and Lovable for behavior-analytic applications.
NLPA (Natural Language Processing for Behavior Analysis): Leveraging Sentiment Analysis, Topic Modeling, and Linguistic Patterns (Kumar et al., 2024) to identify temporal shifts in client or caregiver communication.
3. Human-Centric Supervision & Trauma-Informed Care
Technology should serve the human element, not replace it. Sessions on "The Supervisor Glow-Up" highlighted how we can integrate AI into supervision while staying grounded in:
The ACES Study & ACT: Focusing on Self as Context, Values, and Committed Action.
NARM: Utilizing the Neuroaffective Relational Model for trauma-informed supervision.
4. Impacting Underserved Communities
Finally, I looked at how behavior-analytic programs are being evaluated for system-involved populations, such as the SAFY Initiative and FAIR Training for foster care. Using Behavior Skills Training (BST) at scale ensures that high-quality care reaches the most vulnerable populations.
Final Thought: The future of ABA is data-driven, but it must be ethically grounded. Whether it is correcting statistical skewness in research or building the next generation of clinical "co-pilots," our focus must remain on expertise, pace, and validation.