AAO Position Paper Makes Orthodontist the ‘Human-in-Command’ for Clinical AI

Edited by Orthotown staff

The American Association of Orthodontists has published a position paper setting out six core principles for the use of artificial intelligence in orthodontic care. It designates the licensed orthodontist as the “human-in-command” who remains accountable for every AI tool used in patient care. The paper appears in the September issue of the American Journal of Orthodontics and Dentofacial Orthopedics.

The AAO Task Force on Artificial Intelligence in Orthodontics wrote the paper, adapting it from a version distributed to AAO members in January 2026. The task force describes it as professional guidance rather than regulation.

The six principles are:

• AI governance and the human-in-command

• Regulatory alignment and risk-based oversight

• Trustworthiness and transparency across the product lifecycle

• Patient autonomy

• Education and clinical AI competency

• Operational integration and data privacy

The paper ties oversight to what a software function does and the risk it poses to patients, not to the type of AI behind it. Under the 21st Century Cures Act, clinical decision support functions that meet four exemption criteria fall outside the FDA’s device definition, leaving the orthodontist responsible for their use under the standard of care and state licensure. Functions that direct clinical action without independent review by the clinician may be regulated by the FDA as software as a medical device.

The task force offers orthodontic examples of each:

• A cephalometric function that calculates values from landmarks the clinician places is non-device decision support

• Digital bonding software that designs a transfer tray is regulated as a Class II device

• FDA-authorized image analyzers used for remote treatment monitoring are an example of AI and machine learning software regulated as a medical device

Among its calls to action, the AAO says state dental laws should explicitly prohibit AI technologies that diagnose, prescribe, or manage orthodontic treatment outside a licensed provider’s supervision. It also says AI use should follow the same delegation rules that govern human auxiliaries. Developers are asked to disclose training data, validation methods, and known limitations, and to notify clinicians of algorithm or dataset changes that could affect performance, not only the changes reported to regulators.

The paper also addresses group practice. In large groups and dental service organizations, AI procurement and policy may be handled at the corporate level. The paper says the treating orthodontist still retains professional responsibility for clinical decisions and may unknowingly take on liability for AI deployment choices made above the practice level.

According to the paper, “the orthodontist serves as the human-in-command at the point of care.” It calls comprehensive AI education essential for orthodontists to implement and supervise AI tools and maintain legal and ethical accountability.

Task force members Heather Stone Hopkins, Jonas Bianchi, Lucia Cevidanes, Anthony Puntillo, Mario Tai, Ryan Goy, Frank Hemm, Zelko Relic, Lisa Alvetro, Mindi Brothers, Trista Felty, and Jae Hyun Park authored the paper, which is open access.

Sources:
American Journal of Orthodontics and Dentofacial Orthopedics, “Responsible integration of artificial intelligence in clinical care: An American Association of Orthodontists position paper,” September 2026, Vol. 170, No. 3, pp. 299–321.e4: sciencedirect.com/science/article/pii/S0889540626002416
DOI: doi.org/10.1016/j.ajodo.2026.04.008