Posted: October 3, 2026
October 2026
Healthcare Litigation & Regulatory Monitoring
American Academy of Artificial Intelligence in Dentistry®
Two FDA records — one new, one from last year — together give a useful picture of AI oversight in healthcare, with relevance to dentistry.
In August 2026, FDA opened a forward-looking discussion on how to regulate generative and agentic AI in medical devices. A 2025 case in FDA’s adverse-event database shows the real-world problem already in the system — and why FDA’s questions matter now.
1. FDA opens discussion on generative and agentic AI
On August 18, 2026, the FDA’s Center for Devices and Radiological Health released Considerations for the Regulation of Generative AI-Enabled Medical Devices: Discussion Paper and Request for Feedback.
The document is not guidance and does not establish new regulatory requirements. FDA is seeking public input — comments are due October 19, 2026, under docket FDA-2026-N-7874 — on how its existing medical-device framework should address generative AI-enabled devices. The paper poses 26 discussion questions.
The paper specifically discusses agentic AI — systems capable of autonomously planning and executing multiple steps toward a goal, including interacting with external tools.
Among the issues FDA raises are:
- a two-axis approach to risk: the function the AI performs (from non-directive support to fully autonomous action), weighed against the consequences of relying on an incorrect output;
- whether some generative AI systems need a competency-based approach to premarket evaluation — benchmarking plus clinical confirmation — because every possible input and output cannot be tested;
- how performance should be monitored after deployment;
- how changes to models and system behavior should be managed; and
- how agentic systems and their interactions with external tools fit within the regulatory framework.
Notably, the paper does not address whether these approaches fall within FDA’s existing legal authority or would require new authority from Congress.
These questions will matter as dental AI moves beyond individual applications — image analysis, documentation, patient communication — toward systems that coordinate multiple tasks.
Agentic AI in dentistry was addressed in the Academy’s 2026 peer-reviewed paper, “Agentic Artificial Intelligence in Dentistry: A Practice-Focused Perspective for the Dental Team,” and more recently in its white paper, AI Agents in the Dental Practice: A HIPAA Compliance Guide.
2. FDA database includes a voluntary dental AI report
A separate record in FDA’s Manufacturer and User Facility Device Experience (MAUDE) database concerns an FDA-cleared dental AI system.
The reported event occurred on March 8, 2025, and FDA received the report on March 13, 2025.
According to the voluntary patient report, the AI system identified suspected caries and a periapical radiolucency on dental radiographs. The patient alleged that endodontic treatment was recommended before the dentist personally reviewed the radiographs.
According to the report, the patient then asked the dentist to review the images. The patient stated that after reviewing them, the dentist determined that the suspected finding did not warrant the proposed root canal treatment.
The report also raised concerns about how AI-generated findings were handled within the integrated software environment.
Importantly, this was a voluntary patient report classified as a product-problem report. It does not represent an FDA determination that the device malfunctioned, that inappropriate treatment occurred, or that the clinician or manufacturer was responsible for harm.
The system described in this report was not agentic AI.
That is precisely what makes it relevant to the discussion paper above: the case involves a cleared, conventional AI device — yet it raises the same human-oversight and workflow-integration questions FDA is now asking at a larger scale for generative and agentic systems.
In context
The pairing is deliberate. The August 2026 discussion paper asks how FDA should oversee AI capabilities still emerging. The March 2025 MAUDE entry shows an AI-related concern that already entered FDA’s existing postmarket reporting system — involving a cleared device that is not agentic AI. Together they show the two ends of the oversight challenge: the regulatory questions FDA is now posing, and the real-world cases already testing its current system.
Primary Sources
U.S. Food and Drug Administration. Considerations for the Regulation of Generative AI-Enabled Medical Devices: Discussion Paper and Request for Feedback. August 18, 2026. Docket FDA-2026-N-7874. Comments due October 19, 2026.
U.S. Food and Drug Administration. Manufacturer and User Facility Device Experience (MAUDE), MDR Report Key 21593974, Report No. MW5167620. Event date March 8, 2025; received March 13, 2025.
Related Academy Publications
Farooqi OA, DeNucci DJ, Moharrami M. Agentic Artificial Intelligence in Dentistry: A Practice-Focused Perspective for the Dental Team. Eur J Gen Dent. 2026.
American Academy of Artificial Intelligence in Dentistry®. AI Agents in the Dental Practice: A HIPAA Compliance Guide. 2026.
This brief is provided for educational purposes and does not constitute legal advice. FDA discussion papers do not establish regulatory requirements. Information in MAUDE reports does not, by itself, establish device failure, causation, negligence, or regulatory violation.