The Rise of AI in Nutrition: Navigating Use, Communication & Ethics

Funding from non-CPE revenue for CPE planning, development, review, and / or presentation has been provided by PepsiCo. The speakers received compensation from PepsiCo for their participation. Any opinions or scientific interpretations expressed in this presentation are those of the speakers and do not necessarily reflect the position or policy of PepsiCo, Inc.

The Rise of AI in Nutrition: Navigating Use, Communication & Ethics awards 1.25 CPEUs in accordance with the Commission on Dietetic Registration's CPEU Prior Approval Program.

Q&A

During the live webinar, the speakers received so many great questions—more than we had time to address. After watching the webinar recording above, check out these additional insights from our speakers in the bonus Q&A below.

Q: What are your thoughts related to the use of AI and impact on the environment?
 

Drew Hemler: AI use does in fact have environmental implications, particularly related to energy demand, water use for data center cooling, hardware production, and the broader infrastructure required to train and operate AI systems. As AI becomes more embedded in everyday tools and workflows, sustainability must be part of the conversation. Not as an afterthought but as one dimension of responsible AI use which is included in the BEASTIE framework. 

For nutrition and health professionals, I find the practical question to be: does this use of AI meaningfully improve the quality, efficiency, accessibility, or safety of the work being done? Using AI to support thoughtful planning, improve communication, or reduce administrative burden may be reasonable. Using it casually, repeatedly, or without clear purpose deserves more scrutiny. 

Q: How do I stay on top of AI legal developments?

Barbara Zabawa: The most up-to-date resource with regard to state laws is the National Conference on State Legislatures Artificial Intelligence Legislation Database. This database allows you to filter by topic (such as “health use” legislation and artificial intelligence). The database shows all legislation, not necessarily just bills that have been signed into law. As a result, the database shows you what types of issues each state finds important regarding artificial intelligence regulation.

A good way to stay on top of all legal developments, including federal law and case law developments, I suggest creating a Google Alert. Go to https://www.google.com/alerts and in the search bar, type in “artificial intelligence legal news.” That should give you frequent updates on any legal developments involving artificial intelligence.

Q: Any advice on how to break into the field of Nutrition & AI and informatics? Would love to get involved!

Drew: A great place to start is by building fluency in both directions: learn enough about AI and informatics to understand the language, limitations, and opportunities, while also staying grounded in nutrition practice, ethics, evidence, and patient or population needs. You do not need to become a data scientist overnight. I would start by following credible sources such as the American Medical Informatics Association and the World Health Organization’s digital health and AI guidance.

I also recommend looking for small, practical ways to contribute before waiting for a formal “AI role” to appear. Join interdisciplinary projects, volunteer for technology or quality-improvement initiatives, attend webinars, and document examples of how you are evaluating or improving AI-supported workflows. The strongest entry point is often not “I know AI,” but “I understand nutrition practice and can help translate emerging technology into responsible, useful, human-centered applications.”
 

Q: Which AI tools do you consider as the best for health care providers?
 

Barbara: As discussed in the webinar, provider-facing tools, such as Healthie and Practice Better, that have curated the information that the AI tool can use to provide outputs are the best because you can trust that the information is accurate and most up to date. When practicing in a scientific, evidence-based field, it is important from a standard of care perspective that any outputs from the AI are trustworthy and reliable.
 

Q: What is a “black box” [AI] tool?
 

Drew: A “black box” tool is a system where we can see the input and the output, but we cannot fully see or understand how the tool arrived at its answer. In AI, this can happen when a model produces a recommendation, prediction, summary, or classification without clearly showing the reasoning, data sources, weighting, or decision process behind it.

This matters in nutrition and health because professionals need to evaluate whether an output is accurate, appropriate, equitable, and safe before using it. A black box tool is not automatically “bad,” but it does require caution. The less transparent a tool is, the more important it becomes to verify outputs, understand its intended use and limitations, protect privacy, and keep human judgment at the center of decision-making.
 

Q: Are there any AI platforms that are HIPAA compliant for use in clinical practices?
 

Barbara: While I do not want to endorse any specific company, when looking for HIPAA compliant platforms, keep in mind that HIPAA compliance has several different layers. One layer is that the platform is willing to sign a Business Associate Agreement (BAA) with you. If so, that means that the platform is willing to comply with the HIPAA security rule and inform you if there is a breach of Protected Health Information (PHI). But, you, as the presumed HIPAA covered entity, must also make
sure you have HIPAA-compliant privacy and security practices in place. Having a HIPAA-compliant platform will not absolve you of your own HIPAA compliance responsibilities.

As mentioned in the webinar, most free and low-cost versions of consumer-facing AI platforms are not HIPAA compliant. You must subscribe to the enterprise-level versions to get the company to sign a BAA with you.

Q: How does one protect their website content, book content, article content, etc., from their information being pirated?
 

Barbara: From a legal perspective, the best course of action is to register your creative works with the U.S. Copyright Office. Your original written content (books, articles, blog posts) is automatically copyrighted upon creation but registering it with the U.S. Copyright Office gives you the legal standing to sue for damages. Once your work is published and registered, copyright protection applies normally; that is, others cannot reproduce, distribute, or create derivative works without your permission.

From a technological perspective, my understanding is that you can add protections to your content on the Internet by adding directives within your robots.txt file. I suggest watching some tutorials on how to do this, but once you add these technical protections, AI chatbots will avoid web pages that you have instructed are not to be accessed.