Against the backdrop of genuine AI’s transformative potential, the orthodontic software market has developed a troubling habit of co-opting the language of artificial intelligence to describe conventional software automation. This practice, known as “AI washing,” takes several forms.
The most common variant is the segmentation swap: a platform markets its auto-segmentation capability as “AI treatment planning” when segmentation is the only AI-powered step in an otherwise manual workflow. Six of the ten platforms offer some form of tooth segmentation automation, but only one extends AI to autonomous plan generation. Segmentation is a well-solved computer vision problem. Treatment planning requires understanding orthodontic biomechanics, force systems, and clinical outcomes—a challenge most vendors have not attempted.
Another variant is the rule-based rebrand. OnyxCeph explicitly describes its technology as using “rule-based control modules” and “CA treatment philosophy for automatic transposition.” This is expert-system logic, not machine learning. The system follows encoded rules written by human experts; it does not learn from data or improve with experience. There is nothing wrong with rule-based systems—they can be reliable and transparent—but presenting them as AI is misleading.
A third variant is the speed implication: vendors equate fast workflows with AI intelligence. OrthoUp achieves its approximately fifteen-minute planning time through excellent workflow engineering. The speed is real and valuable, but it comes from human-factor design, not from machine learning.
