The rise of AI tools has made nutrition advice more accessible than ever. Williston students are starting to take advantage of it, but convenience also comes with risks.
Students at Williston are increasingly using platforms like ChatGPT to guide their eating habits, from weight gain strategies to healthier snack choices. Although these tools are convenient and provide personalized plans and strategies,
AI-generated nutrition advice can be unbalanced or misleading, especially for teenagers.
AI is becoming nearly ubiquitous in everyday life, not just in school. It’s providing advice on people’s health and well-being. But AI and accurate facts aren’t always aligned.
A March 2026 study in Frontiers in Nutrition, found that meal plans made by AI for teens often don’t meet proper nutritional standards because they cut too many calories and don’t take into account each person’s needs. Experts say that AI doesn’t have the context or expertise of trained professionals, so following its advice without thinking about it could be dangerous.
Jumpei Ro, a boarding student from Japan, said he turned to ChatGPT for advice on gaining weight and using weight gainer supplements.
“I wanted to put on weight faster, so I asked it what to eat and how weight gainers work,” he said. “It gave me a full plan and explained the effects, which was helpful, but I wasn’t sure if it applied exactly to me.”
He is aiming to gain weight so he can compete more effectively in water polo. Adding size and strength would help him better handle the physical demands of the sport.
The plan included increasing daily calorie intake through foods like rice, eggs, chicken and protein shakes, while suggesting weight gainer supplements as a way to add extra calories quickly between meals. It also explained that weight gainers may lead to excess fat gain, which is exactly what he wanted.
Brooks Marshall, a boarding student from Beverly, Mass., said he used ChatGPT for advice to improve his daily eating habits, especially when it came to snacks.
“After LAS I’d still be hungry, so I’d ask for healthier snack ideas,” he said. “It would suggest stuff that actually kept me full longer, which I liked.”
Brooks, a member of the Varsity Water Polo and Swimming Team, said ChatGPT ensures that he is able to get the correct amount of nutrition to be able to compete at a higher level. ChatGPT gave him suggestions including high protein and fiber, such as Greek yogurt with fruit, peanut butter with apples, protein bars and nuts, all aimed at helping him stay full without relying on more processed options.
To compare the responses of Williston students, I asked questions to Google’s Gemini.
For weight gain, it suggested “increasing caloric intake,” ensuring balanced meals with protein, carbs, and healthy fats, and emphasized consistency over rapid change.
When asked more extreme questions, Gemini’s response became even more restrictive. In some cases, it refused to answer entirely. For example, when prompted with a dangerous dieting request, Gemini responded, “I always try to be helpful, but my safety checks are stopping me from answering this one. If there’s something else you had in mind, I can try that instead.”
Furthermore, when I asked Gemini to build a diet under 1,000 calories per day, it responded, “I can’t answer this one because my safety filters are kicking in. If you have a different question in mind, I’m ready.”
All AI systems, including ChatGPT and Gemini, follow strict codes of conduct that shape what they can and cannot respond to.
“The code of conduct and safety policies for AI on Google Search, powered by the Gemini family of models, forbid the creation of harmful, illegal, sexually explicit, or violent content,” Google wrote in a press release. “These policies prioritize safety, fairness, and the prevention of real-world harm. They aim to ensure responsible AI use by preventing harassment, dangerous activities, and the spread of malicious misinformation.”
