Artificial Intelligence (AI) is everywhere right now, and school nutrition teams are all asking the same questions: Can it build a compliant menu? Can it write standardized recipes that credit correctly? Can it really save time, or does it introduce new risks?

In Health-e Pro’s AI in School Nutrition webinar, Jason Winn, Health-e Pro’s Director of Engineering, covered what AI is good at today, where it falls short, and how to use it in school nutrition programs when compliance is on the line.

The biggest takeaway was simple: AI can be a helpful assistant, but humans should always have the final review. Use AI to support your work and brainstorm, not replace the expertise and verification school nutrition programs require.

What is AI and How Does it Work?

In simple terms, AI is the technology behind things we used to think only humans could do, like understanding language, writing documents, recognizing images, and making decisions.

Most tools we use today don’t “think” the way humans do. AI learns patterns from data to predict what comes next. It generates responses based on probabilities, which is why it can sound confident even when it’s confidently wrong.

Jason explained that most of the AI people use today is text-based, meaning it’s usually good at understanding words, but not great at math or calculations. When text is involved, AI is pretty good at predicting possibilities, but when numbers are involved, it can be wrong.

It can also help with drafting or brainstorming, but it’s risky if you assume the outputs are accurate all the time. That’s why understanding how AI works matters.

The concept of AI has been around for 70+ years. We’ve been using it for a long time already, without even realizing it. Email spam filters, streaming recommendations, and phone autocorrect are all examples of AI that many of us use on a regular basis.

But now, AI is suddenly everywhere. Jason described it as a “perfect storm.” With powerful computers and more data than ever, AI tools are getting smarter and easier for anyone to use.

Where AI is Great Today (and How it Can Help)

Jason shared that AI is helpful when it’s used for low-risk tasks. For example, it can help draft and rewrite parent letters, handouts, and SOPs. It can also summarize long text into a quick overview, which is useful for notes, emails, policy updates, and more.

AI is also great at brainstorming. If you need theme day ideas, marketing copy, or creative ways to promote a menu item or increase participation, it can create ideas in seconds. It can even help you find public information faster, translate language, and format data into something easier to use.

The key is using AI where the risk is low and a human review is easy. It’s great for drafting an email, but it can be risky when you’re trying to identify allergens, where student safety depends on accuracy.

Where AI Tools Can Struggle (and Generate Risk)

One of the main messages of the webinar was that AI can’t do all the work for you, especially when compliance is on the line. Jason highlighted a few common areas where AI tools tend to struggle.

AI tools can be unreliable with math and calculations, current events, and reasoning and logic. They may also state incorrect facts confidently (they “hallucinate”!), and they can mimic expertise without even understanding the topic. And while we’d like to believe AI is unbiased, these tools can pick up bias from the human content they were trained on.

This is why AI can be risky for things like meal pattern compliance, crediting, allergens, serving sizes, and nutrient calculations. It’s not that you can’t use AI to assist with these tasks, but you should always have a human verify anything that may impact compliance or student safety.

Expert Tips for Using AI

Jason emphasized that better AI outputs usually come from better prompts. The more specific you are, the more useful the response tends to be.

Here are a few tips on how to use AI to get better results:
  • Be specific and give context.
  • Tell it the format you want (bullets, checklist, short summary).
  • Talk to it like a human, not a search engine like Google.
  • Assign a role (ex: “write this as if a parent was reading this letter”).
  • Ask it to explain sources, assumptions, or show its work.
  • Verify anything important with a human review.

What’s Next for AI in School Nutrition?

Predicting the future is tough, especially with how quickly AI tools are advancing. But one thing we do know is that they’ll keep evolving. Over time, we’ll see improved transparency, personalization, and compliance. For now, the most realistic way to use AI in school nutrition is as an assistant, not a replacement or autopilot.

In Health-e Pro, you’ll see more purpose-built AI tools being incorporated into our software, with compliance guidelines and standards already in place. We won’t rely on predictions and expect you to figure out if they’re correct or not, because we know that accuracy matters more than creativity when it comes to menu planning, compliance, and student safety.

Key Takeaways

  • AI is a great assistant, but it doesn’t replace human verification.
  • AI generates predictions, not guaranteed facts.
  • The safest use cases are low-risk tasks like drafting, summarizing, and brainstorming.
  • High-risk tasks like compliance, crediting, and allergens still require human review.
  • AI can “hallucinate,” meaning it may provide wrong information confidently.
  • Better prompts lead to better outputs.
  • AI will keep improving, but for now it’s best used as support, not autopilot.

Watch the full AI in School Nutrition webinar and access the resources here!