Today’s Dietitian
Vol. 27 No. 9 P. 28
The field of AI has existed since the 1950s, when scientists first posed the question, “Can machines think?” In the decades since, AI technology has quietly powered many of our everyday tools, from filtering our spam emails to generating product recommendations based on our past purchases. In 2022, the landscape shifted dramatically with the release of generative AI tools like ChatGPT, which made generative AI tools readily accessible to the public. Thus, AI adoption accelerated across industries, with many companies incorporating AI in some way to boost revenue, propel workflow, or improve customer service. The fields of health care and dietetics are no different.
Dietitians are feeling the effects of AI adoption in their work settings. Hospitals are exploring ways to implement AI-Powered Clinical Decision Support Systems (CDSS) to help clinicians make better, safer, and more informed medical decisions.1,2 AI notetakers are integrated into telehealth platforms, saving clinicians time with charting. It may seem like dietitians don’t have a role in a company’s AI implementation, but what happens when the CDSS forgets to account for an individual’s dietary restrictions or cultural foods as it makes diet recommendations? For an AI notetaker application, who is the best person to advise the developers on how to tailor the note structure to dietitians? RDs must step up to help shape the development and implementation of AI tools in health care. They bring expertise in nutrition science, behavior change, chronic disease management, and culturally appropriate care—all of which are critical to designing ethical, accurate, and inclusive AI tools. Without RD involvement, AI risks spreading nutrition misinformation, perpetuating bias, or oversimplifying complex nutrition care.
Dietitians are not just end-users of AI tools. They can be cocreators, leaders, and subject matter experts on AI projects. In this article, Today’s Dietitian interviewed four trailblazing RDs who are at the forefront of shaping AI in the field. Each has their own story about how they came to be leaders at the intersection of AI and nutrition, and urge other dietitians and health professionals to take part in shaping what AI looks like in nutrition care.
Janice MacLeod: Shaping the Future of Diabetes Technology
Janice MacLeod is a diabetes-cardiometabolic consultant and key opinion leader in digital health and chronic condition management. She built her career on a strong foundation in diabetes clinical care before transitioning to work with diabetes technology companies, where she uses her years of expertise in diabetes care to advance AI-powered digital health tools for the diabetes patient population.
MacLeod became passionate about educating dietitians about AI when she came across a discussion board thread of RDs expressing fear about AI taking over their jobs. She stepped into this discussion with a voice of reason and logic, to convey the message that AI is not taking RD jobs and can even make these jobs more efficient if RDs proactively take measures to learn about AI and help shape its implementation. She understood that if the profession failed to get past the unproductive fear and start leading the charge on AI in the field of nutrition, dietitians would be left behind in the AI revolution. Since then, MacLeod started spreading the message through webinars and in-person speaking events, educating thousands of dietitians on the importance of AI literacy and AI leadership. She says, “AI is here, ready or not. We have to figure out how to step up and lead wisely when bringing it into nutrition and health care.”
In the diabetes world, AI is already shaping patient care. Computer vision technology is being implemented into smartphone apps to help patients estimate carbohydrate content in meals and snacks. This technology allows patients to take a photo of their food and receive an accurate macro and calorie count. Studies are showing that the accuracy of computer-vision nutrition estimates is similar to human dietitian estimates.3,4 Companies like Rx Food are already rolling out this technology at scale.5
Automated insulin dosing (AID) systems are advancing fully autonomous insulin dosing with the iLet Bionic Pancreas, Beta Bionics, Irvine, California.6 This device pairs with CGMs like the Dexcom or Freestyle Libre and uses algorithms to automatically adapt insulin doses according to the individual’s changing needs, removing the burden of manually adjusting insulin pump settings.7 Increasingly, AI will be applied to AID algorithms. Imagine pairing AID technology with the aforementioned computer vision technology. Patients take photos of their meals, giving the AID system more data, thus delivering more accurate insulin dosing.
From prevention to detection, to management and treatment, AI is helping dietitians close gaps in diabetes care. Diabetes technology is quickly becoming revolutionized with the power of AI, and we are only at the beginning. Janice urges RDs to stay current, get involved, and be ready to point out red flags when AI tools are inaccurate or unsafe for the populations we work with.
Drew Hemler: Advocating for Health Professionals in Shaping Generative AI
Drew Hemler’s work spans multiple disciplines across dietetics, including teaching at Buffalo State University, providing nutrition care through his private practice, consulting for food companies, and programming for community health and corporate wellness. His work with AI started before the AI boom of 2022, when he was consulting as a nutrition subject matter expert for MSN Health, a Microsoft platform where users can ask health questions and get answers from verified health professionals. Then came the boom of generative AI in 2022, which allowed companies to offload this type of work from humans to AI. But because he has expertise as a dietitian and experience using generative AI tools, Hemler saw the pitfalls of allowing AI to generate health information without oversight from human health professionals. So, he proactively called a meeting with the team overseeing the project and advocated for the project to keep health professionals on the team to advise about what data sets are best for AI retrieval to generate accurate and inclusive answers, how to best communicate health information, and put appropriate guardrails in place.
In addition to his AI consulting work, Hemler speaks about AI in nutrition at nutrition and health conferences globally. He teaches dietitians and health professionals about the current state of AI in health care, the pitfalls of AI, how to engineer prompts, and the ethical considerations of using AI in health care. His overarching message is that dietitians need to actively engage with AI rather than avoid it. This means developing AI literacy and acumen. You do not need to know how to code or develop AI tools to help shape the future of AI.
Through his speaking experience teaching dietitians about AI, Hemler has developed a helpful framework of things to consider when using AI tools. The framework is an acronym called B.E.A.S.T.I.E., and it stands for bias, explainability, accountability, security, transparency, interoperability, and environmental impact.8
The consideration of “bias” means asking what data AI is pulling from and whether it may underrepresent or misrepresent any groups. The data used to train AI models affects its output, and if the data is biased going in, the output will be biased.
“Explainability” describes the ability of the AI system to explain how it arrived at a conclusion. In health care, many things factor into decision-making, and there needs to be transparency about why a certain recommendation is chosen. When using AI tools, the user takes “accountability” for its outputs. The responsibility is on the user to fact-check and edit the outputs as appropriate. “Security” describes how the AI system uses, shares, protects, and stores information. In health care, we must make sure the AI tools used are HIPAA compliant. “Transparency” describes the disclosure of using AI to generate information. Do patients know that AI was used to generate their care plan or write their chart note? Do readers of your blog know how AI was used in the writing process? “Interoperability” is the AI’s ability to work with or integrate with other existing systems. And finally, “environmental impact” considers how using AI tools may affect the environment because AI systems rely on physical data centers that use a lot of energy and water to power the infrastructure. You can read more about the B.E.A.S.T.I.E. framework in Hemler’s article (link in chart).
Hannah Kittrell: Leveraging the Predictive Power of AI to Shape Precision Nutrition
Hannah Kittrell is a dietitian and PhD candidate specializing in biomedical AI. Her interest in precision nutrition and using data to predict how a person will respond to specific foods or dietary patterns led her to pursue a doctorate. Because studying precision nutrition requires analyzing vast amounts of complex data, she uses machine learning algorithms to explore patterns in data and make predictions.
To understand precision nutrition and the role AI can play in this area of research, consider the field of the gut microbiome as an example. The gut microbiome is extremely complex to study, with trillions of diverse microorganisms that exist within a cocktail of metabolites and biochemicals that are dynamic and interactive. Studying the gut microbiome requires collecting large amounts of data and analyzing it. AI is really good at analyzing large datasets and drawing out associations.
Kittrell uses the predictive and analytical power of AI to process large volumes of nutrition data and uncover patterns—not the typical text-generating large language models (LLMs) most people think of when they think of AI. Through her PhD program, she has learned coding languages such as Python, R, and SQL, which she uses to build and tweak machine learning models before applying those models to datasets. Much of her daily work involves cleaning data and checking for bias, since the quality of the AI’s output depends directly on the quality of the input. For example, if a dataset of metabolic marker responses to a specific diet intervention underrepresents women or certain ethnic groups, the model may draw conclusions that only reflect the responses of the majority population, masking important differences in how diverse groups actually respond to the same dietary intervention. As the only dietitian on her team, Kittrell brings her expertise to make sure data is representative of real-world patient populations and offers nutrition context when analyzing and interpreting data.
Kittrell views AI as a helpful tool that can accelerate RD work if it is used wisely and responsibly. For example, some of the data crunching that she does could be done with traditional statistical models, but using AI and machine learning algorithms allows her to process much more complex data faster. In one of her main projects, Kittrell is using machine learning to evaluate dietary intake patterns in 100,000 individuals across hundreds of different foods—a far more daunting task if using only traditional statistics without AI. Still, Kittrell knows not to rely solely on AI interpretations of data and applies her own clinical judgement to make sure the data is non-biased, makes sense, and can be applied in appropriate contexts.
Raul Palacios: Teaching the Next Generation of AI-Assisted Dietitians
Raul Palacios is a dietitian and director of the DPD program at Texas Tech, where he integrates AI into MNT courses and teaches an elective AI fundamentals course. He is also currently pursuing a PhD in AI in nutrition. His dietetics career has spanned inpatient clinical work, clinical foodservice, and now academia.
One of the classes that Palacios teaches is a very popular elective class called “AI Fundamentals with Health & Human Sciences Applications.” In this class, health professional students learn about how AI works, the applications of AI and machine learning/deep learning, and the ethical implications of AI in the health care setting. Students have opportunities to explore AI in their field of health care and propose a novel AI use case or governing policy related to their field.9
In his MNT class, the curriculum is not specifically about AI, but AI tools are used to facilitate student learning. For example, students use a custom AI chatbot to simulate a diabetes education session. Students interact with the chatbot as though they are speaking to a patient with diabetes. The chatbot then gives feedback on their nutrition education and counseling skills. Students also have access to custom-built AI chatbots to practice job interviews and receive resume and cover letter guidance.
Palacios believes that students must first learn to do things manually without AI assistance, but he also sees an important role for AI in student learning because students will inevitably encounter these tools in the workforce and need to be prepared. He compares this to teaching students how to calculate resting metabolic rate with the Mifflin-St Jeor equation before using an Excel sheet calculator for automatic calculation. By learning how weight, height, and age interact in the formula, they are better equipped to recognize errors, such as accidentally using inches instead of centimeters for height. In the same way, students who first build a strong foundation in nutrition knowledge can use AI more effectively and responsibly, since they are able to recognize when the tool makes a mistake or overlooks key information.
But not all professors are AI-friendly. Some prohibit students from using it altogether. Palacios says he sees both sides of the argument; students using AI could potentially hinder their knowledge development if AI is used for doing the grunt work, hindering student learning.
A recent study from MIT10 measured the brain activity of 54 participants writing SAT-level essays. One group of participants was allowed to use an LLM, a second group was allowed to use a search engine without an LLM, and the third group was only allowed to use their brains to write the essays. Three essay-writing sessions were conducted over three months on various SAT essay topics. Brain activity was the highest in the brain-only group and lowest in the LLM group. The group using the LLM to write their essay also displayed a weak ability to remember points in their essays. The researchers then conducted a fourth crossover session where they assigned the LLM group to brain-only and the brain-only group to LLM. The LLM-to-brain group consistently showed weaker neural connectivity and performed only slightly better than the brain group’s first essay-writing session, suggesting that the LLM group would need to start over from the beginning if they wanted to achieve similar results as the brain-only group by the third session. In the brain-to-LLM group (participants who had three sessions of essay-writing using only their brain, then did the fourth session with an LLM), the prompts they developed were more specific in seeking certain elements of information rather than simply entering, “hey Chat, write this essay for me,” suggesting more critical thinking. The authors note one concerning finding: in the fourth session, when the LLM group was assigned to write the essay using only their brains, the essay topics showed a narrower set of ideas. However, the sample size of this study is a limitation to keep in mind. By the fourth session, the number of total participants had dwindled from 54 to 18. Nevertheless, how do we teach students in a world of easily accessible AI tools while still making sure they learn foundational knowledge and sharpen their critical thinking skills?
In Conclusion
The unifying theme across all interviews shared here is that dietitians must proactively insert themselves at the table of AI decision-making in their organizations. None of the interviewees had a blueprint or job description for using, teaching, or influencing AI implementation within their organizations. MacLeod was inspired to lead after witnessing unproductive fear dominate conversations about AI in nutrition. Hemler advocated for dietitians and health professionals to remain in the process of communicating and disseminating health information after his company incorporated generative AI. As the only dietitian in her PhD program, Kittrell highlights the critical perspective dietitians bring to analyzing and interpreting biomedical data. Palacios is pioneering a path for educators to prepare the next generation of health care professionals to work effectively alongside AI. These trailblazers show that RDs can and must help shape the future of AI in health care and nutrition. Now, it is up to all RDs to add their voices and ensure that the profession helps define how AI is used in the field of nutrition.
— Jane Guo, MCN, RD, LD, is the owner of Habits Nutrition Counseling, where she specializes in helping clients achieve sustainable weight loss. In addition, she is a professional speaker and consultant on the integration of artificial intelligence in nutrition, teaching dietitians how to use AI tools to streamline practice, enhance patient care, and shape the future of the profession.
Janice MacLeod, MA, RD, CDCES, FADCES
Author, Speaker, and Diabetes Cardiometabolic Health Consultant
MacLeod is an expert in diabetes and cardiometabolic health. She combines her expertise as a dietitian and her extensive knowledge about AI to consult for diabetes technology companies.
LinkedIn: www.linkedin.com/in/janicemacleod
Hannah Kittrell, MS, RD, CDN, ACSM EP-C
PhD Candidate, Biomedical AI at Mount Sinai
Kittrell uses AI/ML to analyze large datasets to study precision nutrition. As the only dietitian on her team, she ensures that the data is analyzed without bias and is interpreted in appropriate contexts.
LinkedIn: www.linkedin.com/in/hannah-kittrell
Drew Hemler, MSc, RD, CDN, FAND
AI in Health Care Consultant and Speaker
Hemler uses AI tools in his many roles as consultant, speaker, professor, nutrition communicator, business owner, and dietitian. Through his speaking, he teaches other dietitians and health professionals how to responsibly use AI tools.
LinkedIn: www.linkedin.com/in/drewhemler
Raul Palacios, MS, RDN, LD
DPD Director at Texas Tech
Palacios teaches the next generation of dietitians how to work alongside AI. He is also currently pursuing a PhD related to AI in nutrition.
LinkedIn: www.linkedin.com/in/raul-palaciosms-rdn
References
1. Elhaddad M, Hamam S. AI-driven clinical decision support systems: an ongoing pursuit of potential. Cureus. 2024;16(4):e57728.
2. Romero-Brufau S, Wyatt KD, Boyum P, Mickelson M, Moore M, Cognetta-Rieke C. Implementation of artificial intelligence-based clinical decision support to reduce hospital readmissions at a regional hospital. Appl Clin Inform. 2020;11(4):570-577.
3. Vasiloglou MF, Mougiakakou S, Aubry E, et al. A comparative study on carbohydrate estimation: GoCARB vs. dietitians. Nutrients. 2018;10(6):741.
4. Shonkoff E, Cara KC, Pei XA, et al. AI-based digital image dietary assessment methods compared to humans and ground truth: a systematic review. Ann Med. 2023;55(2):2273497.
5. RxFood website. https://rxfood.com. Accessed August 26, 2025.
6. FDA clears new insulin pump and algorithm-based software to support enhanced automatic insulin delivery. FDA website. https://www.fda.gov/news-events/press-announcements/fda-clears-new-insulin-pump-and-algorithm-based-software-support-enhanced-automatic-insulin-delivery. Published May 19, 2023. Accessed August 26, 2025.
7. Beta Bionics website. https://www.betabionics.com. Accessed June 13, 2025.
8. Hemler D. How dietetic practitioners can responsibly navigate AI: the B.E.A.S.T.I.E. Framework. LinkedIn website. https://www.linkedin.com/pulse/how-dietetic-practitioners-can-responsibly-navigate-drew-zpcbc/?trackingId=0VR6pM4JTxWhitqoMD%2FM6w%3D%3D. Published May 22, 2025.
9. Palacios R. I had a lot of interest in our “AI Fundamentals with Health & Human Sciences Applications” course at the Teaching & Learning With AI Conference last week. LinkedIn website. https://www.linkedin.com/posts/raul-palaciosms-rdn_i-had-a-lot-of-interest-in-our-ai-fundamentals-activity-7335991217479917568–9oU?utm_source=share&utm_medium=member_desktop&rcm=ACoAACDiOd8B2jyfCij25khfgVR_vyC9fzyfrjE. Published July 2025.
10. Kosmyna N, Hauptmann E, Yuan YT, et al. Your brain on ChatGPT: accumulation of cognitive debt when using an AI assistant for essay writing task [published online June 10, 2025]. arXiv. doi: 10.48550/arXiv.2506.08872.
