Artificial intelligence is rapidly changing the way we work, and dietetics is no exception.
From drafting client resources and summarising information to supporting administrative tasks and streamlining workflows, AI tools are becoming increasingly accessible to healthcare professionals. Used well, they have the potential to save time, reduce repetitive work and help us practise more efficiently.
But conversations around AI in healthcare often jump quickly to one question: Will AI replace healthcare professionals?
Personally, I think that often asks the wrong question.
Rather than asking, “How do we remove the practitioner?”, I think a far more useful question is:
“How do we better equip the practitioner?”
AI can be incredibly useful for structuring rough notes, drafting general education materials, simplifying complex information, supporting administrative tasks or helping us explore an unfamiliar topic.
One of the things I value most about AI is its ability to broaden my thinking. Sometimes it raises a consideration I had not thought to explore, gives me terminology for a deeper literature search or sends me down a research rabbit hole I may not otherwise have considered.
That does not mean blindly accepting the answer. It means using AI to expand curiosity, then applying evidence and clinical reasoning to decide what is actually useful. Human expertise is not perfect either; we all have knowledge gaps, limited time and biases, and access to mentors or colleagues is not always immediate.
For me, that is where the real opportunity lies: using AI to complement our expertise, reduce repetitive work and give clinicians more time for the work that genuinely requires it.
After all, we did not complete years of training to spend hours reformatting documents, switching between tabs or repeating administrative tasks that technology can help streamline.
But in healthcare, efficiency cannot come at the expense of privacy, accuracy or clinical reasoning.
So, how can dietitians actually use AI safely?
1. Protect Client Privacy
One of the most important considerations when using AI is protecting client information.
Before entering anything into an AI platform, dietitians should understand how that platform stores, processes and uses data. Names and direct identifiers should be removed wherever possible, unnecessary personal details should be left out, and only the minimum information required for the task should be shared.
A simple question I like to ask is:
“Does this AI tool actually need this information to complete the task?”
If the answer is no, leave it out.
It is also important to consider whether the tool is appropriate for the information being entered in the first place. A general-purpose chatbot may be useful for brainstorming or rewriting text, but that does not automatically make it suitable for handling sensitive health information. Workplace policies, professional obligations and the platform’s own privacy and data-handling practices still apply.
2. Treat AI Output as a Starting Point, Not the Final Answer
As impressive as AI can be, I’m sure many of us have encountered the GPT hallucination: an answer delivered with complete confidence that is outdated, misleading or simply wrong.
No tool is perfect. As healthcare professionals, we are trained to think critically, interpret information and apply clinical judgement. Using AI should be no exception.
This matters particularly in dietetics because recommendations rarely depend on one factor alone. A person’s needs may be shaped by their medical history, medications, pathology, food preferences, culture, accessibility, finances, eating behaviours and broader psychosocial circumstances.
AI can help bring information together, but it does not automatically understand which details matter most for the individual sitting in front of you.
My approach is to treat AI output as something to review, question and refine. If it makes a clinical claim, check it. If it provides a reference, verify that the reference exists and actually supports the claim. If it makes an assumption, ask whether that assumption is justified.
3. Know Where AI Is Most Likely to Get Things Wrong
Some tasks deserve more caution than others.
One area I am particularly careful with is nutrient calculation.
Many dietitians have probably already had clients mention using AI to photograph a meal and estimate its calories or macros. It sounds convenient, but there are several opportunities for error. The tool may misidentify the food, estimate the portion incorrectly, confuse raw and cooked weights, miss oils or sauces, or generate nutrient values that appear precise without a reliable underlying calculation.
From my own experience using general AI tools, I have seen nutrient values hallucinated, foods misidentified and calculations shift when the same question is asked in a slightly different way.
This does not mean AI has no role in nutrition analysis. It means we need to understand what sits underneath the output.
Where did the nutrient data come from? How were ingredients matched? Was a recognised food composition database used? Were portion sizes estimated or actually provided? Can the result be traced and checked?
A number that appears to two decimal places can still be wrong.
The same caution applies to clinical guidelines, research papers and references. Generative AI can invent citations, misquote recommendations, present outdated guidance as current or fill in missing information with plausible-sounding assumptions.
The more clinically significant the output, the more important verification becomes.
4. Remember That AI Can Reflect Bias Too
AI systems are trained on large amounts of existing data, which means they can reproduce the same biases and gaps already present in that information.
In dietetics, this can show up in subtle ways. A “healthy meal plan” may default to Western foods. Weight-related recommendations may be overly simplistic. Suggestions may assume someone has access to a full kitchen, a large grocery budget, high health literacy or particular ingredients.
A meal plan can look nutritionally balanced on paper and still be completely inappropriate for the person receiving it.
This is where dietitians remain essential.
We understand that food is not just nutrients. It is culture, access, family, identity, routine, finances, preferences and lived experience. AI-generated suggestions still need to be assessed within that broader context.
So, What Should Dietitians Ask Before Using AI?
Rather than following a long checklist every time, I think it comes down to a few core questions:
- Am I protecting the client’s privacy?
- Is this the right tool for this task?
- Can I verify the important parts of the output?
- Am I still applying my own clinical reasoning?
- Would I be comfortable taking professional responsibility for the final result?
If the answer to any of these is unclear, the output probably needs more review.
The Future Is Not AI or Dietitians
The conversation around AI in healthcare is often framed as a choice between humans and technology.
But I think that framing misses the point.
The future of dietetics should not be about choosing between AI and dietitians. It should be about using technology thoughtfully, while protecting the parts of care that require human judgement, empathy and clinical expertise.
AI is incredibly powerful. It can expand how we think, accelerate how we work and take repetitive tasks off our plates. But the more powerful the tool, the more important it is to use it responsibly.
As Uncle Ben so wisely told Spiderman, “With great power comes great responsibility.”
Privacy still matters. Evidence still matters. Context still matters. And the clinical reasoning of dietitians remains absolutely essential.
So no, I don’t think the answer is replacing dietitians with AI.
I think it is about giving dietitians better tools, clearing away more of the repetitive work and creating more space for us to do what we actually trained to do: think critically, connect with people and provide care that understands the human being behind the nutrition problem.
Because at the end of the day, food is deeply human.
And dietetics should remain that way.
Emillya T.