A Look at the ADA’s Whitepaper on Artificial Intelligence Uses in Dentistry

The ADA 1106-2022 whitepaper “Overview of Artificial and Augmented Intelligence Uses in Dentistry” provides a comprehensive analysis of the state of artificial intelligence (AI) in dentistry, as of the end of 2022. Although, by March 2025, AI has already advanced, the takeaways are still relevant for oral and maxillofacial surgeons curious about AI use in various dentistry specialties.

Oral and maxillofacial surgery (OMS) professionals and representatives of other dental specialties helped develop the whitepaper as members of the ADA Standards Committee on Dental Informatics. The OMS professionals noted as contributors are Eric Pulver of Denti.AI and Pulver Oral Surgery and Larry Stigall and Sandy Guenther, both of the American Association of Oral and Maxillofacial Surgeons.

Background & Context

The first robotic dental surgery system was cleared by the Food and Drug Administration for dental implant procedures in 2017. By the end of that year, China had developed the world's first autonomous-guided dental implant placement system.

To function effectively, all AI systems require extensive training on large datasets. In the context of dental imaging, for example, the training dataset typically consists of dental images, such as intraoral radiographs. The whitepaper notes that while the “ground truth” or “gold standard classification” for training an AI system would ideally be expert analysis — specifically by oral and maxillofacial radiologists — there is no guarantee of “ground truth” in most specific products or services.

Despite these limitations, AI systems can provide clinicians with vast amounts of information on population health patterns. However, it remains the clinician's responsibility to determine how this information applies to individual patients.

The majority of the whitepaper presents examples and case studies illustrating how various facets and specialties within dentistry are utilizing AI. Below, we summarize a few highlights.

AI Use in the Clinical Setting

First, let’s explore current and potential uses in the clinical setting.

• Dental Implants: By 2022, various AI models had been developed to recognize implant types in periapical and panoramic radiographs. These models are reliable in determining osteointegration success, implant prognosis, and identifying fractured implants. According to one study, AI models' accuracy for implant type recognition ranges from 93.8% to 98%, while implant success prediction varies between 62.4% and 80.5%.
• Oral and Maxillofacial Surgery: The whitepaper notes that AI application in OMS was limited at the time but expected to grow. AI can aid in diagnosing and planning treatment with minimal errors. One study demonstrated AI's ability to predict subgroups of temporomandibular joint internal derangements with high specificity and sensitivity, potentially reducing diagnostic dilemmas. The whitepaper does not extensively describe the role of AI in OMS, and many ACOMS Review readers are likely familiar with other AI use cases within the subspecialty.
• Dental Radiology: AI supports clinical decision trees for selecting appropriate imaging examinations. It can identify radiologic manifestations of disease, such as coronal radiolucencies, periodontal bone loss, apical radiolucency, and mandibular condyle erosion.
• Temporomandibular Joint Disorder (TMD): AI-based nightguard design, using intraoral scanning, is transforming this area. These designs consider opposing occlusion to ensure a proper bite, making the process scalable and repeatable. If a patient loses their nightguard, it can be quickly reprinted or milled and shipped overnight.
• Teledentistry: AI assists in interpreting images collected during remote consultations. Use cases include triaging maxillofacial trauma patients, screening children's dental conditions via smartphone photos, and designing dental smiles using scanned images. These approaches are cost-effective and show acceptable accuracy compared to in-person exams.
• Electronic Dental Records (EDR): AI helps identify unscheduled treatments and missed diagnoses by analyzing complex patient data. EDR/EHR systems store and integrate high-quality data and images, essential for developing AI tools. Vendors like BigMouth support this by providing standardized databases from dental schools.
• Scanning: Digital scanning of teeth and tissues is now common, reducing patient discomfort and improving clinical efficiency. However, challenges remain, such as clinicians' technical proficiency and the quality of imagery, which is crucial for AI accuracy.

AI Use Across the Practice

AI is being used to support dentistry practices in non-clinical ways as well.

• Claim Processing: AI systems are commonly used in the payor claim review process. They validate the type and content of attachments submitted with claims and screen radiographic images to determine if proposed or completed treatments meet payor clinical guidelines. Current AI products can identify the type of radiograph, each tooth within a radiograph, and anatomic structures, existing restorations, and pathologies. If the correct information is not present in the radiograph, the claim may be returned to the dental office with a request for the correct image.
• Payment Integrity: AI processing of radiographs can accurately identify and segment teeth, their associated anatomy, existing restorations, and various pathologies. These representations allow AI to compare submitted radiographs to those previously submitted with other claims. Importantly, AI can detect similar images even when altered by cropping, resizing, contrast adjustment, or other manipulation techniques.
• Quality Assurance: A unique capability in development is the use of AI computer vision to monitor staff performance by comparing dental consultant decision-making to their peers and to gold-standard AI models of performance.

Regulatory Environment

From a regulatory standpoint, one of the most important features to recognize about the use of AI in providing dental care is that it is strictly a supplement to the clinician. Dentists hold the responsibility for diagnosis, prevention, care, and treatment of oral diseases and conditions under the authority of state licensing agencies. Within their scope of practice, dentists use a variety of images, subjective and objective information, and tools.

Globally, practitioners must balance the benefits and risks when assessing AI. While AI is an important and exciting decision support tool, the responsibility for its use, and the associated benefits and risks, ultimately depends on the decision authority of the dental professional using AI. A global voluntary group, the International Medical Device Regulators Forum (IMDRF), provides guidance on medical device regulation and develops internationally agreed-upon guidance for medical devices.

The whitepaper recommends a list of six questions to consider before incorporating a new AI system:

1. What tasks are claimed as part of the system’s intended use?
2. How was it established that the validation dataset had enough images for each classification task claimed as part of intended use?
3. How was it established that the validation dataset had sufficient variety in gender, age, and ethnicity? And how was it established that there were enough images for each subpopulation?
4. Was the validation dataset sequestered from the training and testing processes?
5. If the system’s intended use includes treatment planning, what rate of false positives should be expected?
6. If the system’s intended use includes radiographic screening, what rate of false negatives should be expected?

Since the whitepaper’s publication, in the United States, the ADA has been actively involved in setting standards for AI in dentistry. As of 2025, the ADA has released several standards and technical reports to guide the responsible use of AI in dental practices These standards emphasize the need for independent validation datasets, transparency, and fairness in AI applications.

The whitepaper goes into greater detail on the value of artificial intelligence, including distinguishing between artificial intelligence and artificial augmented intelligence, which emphasizes the involvement of a human. As oral and maxillofacial surgery has benefited for decades from advancing technology, AI expansion is no different. It’s beneficial for OMS professionals to be aware of the environment, analyze options carefully, and incorporate new technology they decide is beneficial for their workflows, staff, and patients.