Artificial intelligence (AI) is now closely entwined with every aspect of modern life. It influences how we communicate with people, manage our money, run our homes, engage with personal interests, and so much more. It also heavily influences business success, whether it is utilised for marketing, financial management, quality assurance or other purposes. The healthcare sector has not escaped the AI revolution either, with new technology driving risk assessments, diagnostics, treatment planning, and more in various medical settings. In dentistry, AI is being used to streamline workflows, elevate the patient experience, and deliver more predictable outcomes. For the best and safest implementation of the technology, it is crucial that we as dental professionals understand how it works, its benefits, and its limitations.
The LLM explained
Large Language Models (LLMs) are neural networks that use deep learning. They access massive datasets to identify patterns and extract information, allowing them to predict future patterns or answer questions. Simpler forms of the technology have been used in the form of predictive text on our mobile phones for decades. More advanced versions, like transformer models, are capable of unsupervised training, using information from billions of web-based sources to learn from.[i] It’s important to realise that LLMs do not ’think’ in the way that humans do; they simply interpret data or deduce outcomes based on the past data available.
Today, LLMs are utilised in a broad range of indications, from copywriting and text clarification to code generation, and more. Different types of models exist, classified according to their primary function. Autoregressive models, for example, generate text – think OpenAI ChatGPT and Meta LLaMA. Masked language models fill in missing words rather than generating content from scratch, while general encoder-decoder models can read texts and produce new versions like translations, summaries, or rewrites. Finally, retrieval-augmented generation models can access search engines or specific databases to retrieve facts that help them respond to questions asked.[ii] Each type of LLM has its own strengths and weaknesses, and appreciating this is the key to using the right platform for specific tasks.
Success stories
In dentistry, AI-driven technologies have been successfully deployed. For example, they have been instrumental in improving the treatment journey and elevating the quality of patient care.[iii]
In endodontics, AI has been shown to improve the accuracy and efficiency of treatment procedures, with more precise diagnostics, enhanced treatment planning, and supported clinical decision-making.[iv] In particular, AI programmes have proven their worth in situations like automated canal morphology detection, caries diagnosis, pulpal condition assessment, and more.
In addition, emerging applications for AI include remote treatment monitoring and teledentistry, which are changing the way dentistry is delivered. Initial evidence suggests these systems to be particularly useful in orthodontics, detecting debonded brackets and assessing oral hygiene as part of ongoing care.[v]
Still learning
Despite its enormous potential to benefit dentistry, AI still comes with its limitations and challenges. When using its capabilities to care for patients, it is essential that dental professionals understand what these are and how to manage them.
The first is a question of ethics, which is a highly complex subject. AI-related ethical issues can be split into three main categories – epistemic (when dealing with misguided, inconclusive or inscrutable evidence); normative (when managing unfair outcomes); and concerns about traceability.[vi] Challenges around technical reliability, accountability, and data security are often cited as barriers to AI adoption by medical facilities.[vii] This emphasises the need for meticulous regulations, monitoring, and user training, ensuring that AI-assisted technologies are implemented in the safest, most effective, and compliant way.
I was also interested to discover that not all LLM platforms are equal. A comparative study published in early 2026 found that, while most of the 11 evaluated LLMs demonstrated high accuracy and reproducibility, they still varied significantly in performance.[viii] The authors highlighted that the technology continues to require strict validation and expert oversight to support safe and effective clinical intervention. They tested the models with endodontic-related questions and noted considerable progress, though further refinement is needed before the technology can be confidently relied upon.
Stay safe
There is clearly a place for AI in modern dentistry, with its ever-evolving applications continuing to transform the delivery of patient care. For dental practices committed to operating at the cutting-edge, AI-driven technology will be an unavoidable stepping stone to business growth and development. To take advantage of its extraordinary capabilities, it’s vital to have at least a basic understanding of what it does and how it works. Only then can you maintain adequate data protection, patient safety, and clinical quality.
For further information about the endodontic referral services available from EndoCare, please call 020 7224 0999 or visit the NEW website www.endocare.co.uk
Author: Michael Sultan
[i] What is LLM (Large Language Model)? Amazon. https://aws.amazon.com/what-is/large-language-model/ [Accessed January 2026]
[ii] Ponfoort D. Understanding different types of LLMs: strengths and weaknesses. Ai governance. July 2025. https://www.cleverrepublic.com/resources/blog/understanding-different-types-of-llms-strengths-and-weaknesses/ [Accessed January 206]
[iii] Tyagi M, Jain S, Ranjan M, Hassan S, Prakash N, Kumar D, Kumar A, Singh S. Artificial Intelligence Tools in Dentistry: A Systematic Review on Their Application and Outcomes. Cureus. 2025 May 29;17(5):e85062. doi: 10.7759/cureus.85062. PMID: 40585609; PMCID: PMC12206247.
[iv] Asgary S. Artificial Intelligence in Endodontics: A Scoping Review. Iran Endod J. 2024;19(2):85-98. doi: 10.22037/iej.v19i2.44842. PMID: 38577001; PMCID: PMC10988643.
[v] Thornton OR. Artificial Intelligence in Dentistry: A Systematic Review and Meta-Analysis of Techniques, Clinical Applications, and Performance Outcomes. Carolina Digital Repository. University Libraries. October 2025. https://doi.org/10.17615/ze1w-s425
[vi] Morley J, Machado CCV, Burr C, Cowls J, Joshi I, Taddeo M, Floridi L. The ethics of AI in health care: A mapping review. Soc Sci Med. 2020 Sep;260:113172. doi: 10.1016/j.socscimed.2020.113172. Epub 2020 Jul 15. PMID: 32702587.
[vii] Hou J, Cheng X, Liao J, Zhang Z, Wang W. Ethical concerns of AI in healthcare: A systematic review of qualitative studies. Nursing Ethics. 2025;0(0). doi:10.1177/09697330251385024
[viii] de Araújo LP, Moreno LB, de Araújo BCC, Chaves ET, Botero TM, Romero VHD. From Evidence-Based Endodontics to Generative AI: A Comparative Study of Eleven Large Language Models. J Endod. 2026 Jan 21:S0099-2399(26)00010-5. doi: 10.1016/j.joen.2026.01.009. Epub ahead of print. PMID: 41577028.
