Yes, artificial intelligence is increasingly being used to help people manage their diabetes. This can include reminders about medications and prescription refills, to developing an exercise plan or creating shopping lists for specific food choices.

We have reached a point where we can type a command into a prompt, and artificial intelligence will spit out entire paragraphs and instructions in response to the question we’ve asked. This includes for managing diabetes.

You can ask ChatGPT or Google Gemini how to help manage your diabetes each day, from ways to track your medications to ways to lower blood sugar levels, and different exercises or foods to try that might have less effect on your glucose levels.

AI tools are already making a dent in diabetes management, as researchers and the medical community research this trend and determine how to best guide people in using AI to help manage their condition.

In just seconds, AI can generate personalized nd proactive tips, and you can build on that initial information as much as you’d like, digging deeper and personalizing what AI tells you even more.

While convenient and helpful at times, it’s important to remember that AI may not always be accurate and is not a replacement for real-time consultation and care from a healthcare professional.

This may seem like a basic task, but some AI tools are helping people remember and manage aspects of their diabetes care plan. That might include appointment scheduling, reminders about taking medications at a certain time, or ordering more refills at the pharmacy.

In some of the most basic ways, people can enter into software or online tools their specific information for dosing and filling prescriptions, to help generate a calendar or daily reminders that can sync with phone alerts or alarms to help them take a medication.

Other AI-based apps may serve as a “virtual health coach,” providing these type of medication reminders and also tracking progress or sending personalized nudges to enourage healthy diabetes behaviors to help manage the condition.

Technology and smart algorithms

With the recent emergence of AI tools, people with diabetes may also hear AI mentioned in the context of existing diabetes technologies, including continuous glucose monitors (CGMs) and insulin pumps.

These are known as hybrid or fully closed loop systems, where they use data to help automatically adjust insulin doses based on your current needs.

The technology does use smart algorthms that serve as a “brain” for some decision making. These became more common in the past decade. However, they are not on the same level or are the same as newer AI tools generating information and guidance.

Importantly, AI tools are not meant to replace a doctor’s guidance and advice on diabetes management. This includes for personalized insulin dosing and changes to medications or eating patterns. Always consult your diabetes care team before making any changes to your insulin amounts or other aspects of your individual care plan.

AI-powered apps and programs may provide personalized, real-time recommendations for food choices, portion sizes, and carb and nutritional information.

Carb counting and nutritional insights

Some apps can even estimate carbohydrate content from a picture of a meal.

While eating styles and food choices are personal choices that should be discussed and established with your diabetes care team, some people with diabetes may choose to follow a low-carboyhdrate eating style.

Research often cites this eating style as a way for people with diabetes to improve blood sugar levels, as foods with less simple carbohydrates may not dramatically spike blood sugars in the same ways as higher-carb foods.

Carb counting can be a challenging and exhausting endeavor, though. That’s where AI might help. Certain apps and programs may offer different AI options for this:

  • Photo recognition: Simply snap a photo of your meal (e.g., a plate with chicken, avocado, and broccoli). The AI instantly identifies the food items.
  • Portions: Using the image and, sometimes, a reference object (such as a credit card or a ruler) for scale, the AI estimates the portion size.
  • Calculating carbs and nutrients: AI may cross-reference this information with a large nutritional database to provide an accurate carb count estimate, along with protein and fat. This may reduce the guesswork and mental load tied to carb counting.

Personalized meal plans

AI can act as a personal low-carb dietitian, constantly adapting its suggestions to your needs and preferences. This might include generating specific meal plans or recipes for a day, or possibly creating shopping lists to help guide you at the grocery store.

  • Generating low-carb plans: You can input your daily carb goal and the AI will generate full weekly meal plans, including recipes and grocery lists, that fit your target.
  • Adaptive suggestions: If you input an ingredient you have on hand (such as chicken breast and spinach), the AI can instantly suggest a low carb recipe that utilizes those items.
  • Exclusion lists: You can instruct the AI to exclude specific ingredients, such as foods you dislike or are allergic to, ensuring the meal plan is fully customized.

Using AI to help manage gastrointestinal symptoms

An example might be for people with diabetes who develop gastrointestinal complications, meaning they may need to adjust their eating style and food choices to manage this digestive issue in the context of blood sugar management and insulin dosing.

  • Diabetes need: Keep total carbs under X grams.
  • Gastroparesis need: Low-fat, low-fiber, and liquid or soft consistency.
  • Possible triggers: As you log meals and symptoms (like nausea, bloating, or vomiting), the AI uses machine learning to identify your unique food triggers, even within acceptable categories. For example, it might learn that while most soft proteins are fine, a specific artificial sweetener used in a liquid diet supplement consistently leads to bloating for youo, and it will flag that ingredient in future suggestions.
  • Data generation: AI then cross-references massive databases to suggest recipes and meal plans that satisfy all conditions, a feat that is nearly impossible to do manually.

Always make sure to consult your diabetes care team, including a dietitian or nutritionist, for specific help in navigating food choices for your diabetes management if you experience any other health conditions or complications.

AI-powered apps may also provide personalized, real-time recommendations for physical activity.

This can be fine-tuned based on your own perferences for activity intensities or times of day, and even with specific exercises or activities you may enjoy or not want to do.

With everything AI, this is not meant to replace the individualized care from your diabetes care.

Depending on the questions or prompts you offer, the AI tool may offer guidance on before or during activity levels. This could include:

  • “Your glucose is trending down. Eat 15g of fast-acting carbohydrate now to prevent hypoglycemia during your 30-minute walk.”
  • “Start your resistance training now. No pre-exercise snack is needed, but monitor closely for a post-workout rise.”

AI is also helping people better understand their risk for developing diabetes and diagnosis it earlier, preventing delays in diagnosis and care that can lead to dangerous situations.

For people with type 1 diabetes and latent autoimmune diabetes in adults (LADA), this can mean helping to prevent high blood sugar levels and diabetes-related ketoacidosis (DKA), which often develops in the later early stages of T1D and can quickly become life threatening.

People with type 2 diabetes may live with the condition for years, not experiencing many noticeable side effects until complications begin to develop.

Using AI has become a way to prevent diagnosis delays and inform people of their diabetes risk.

Specifically, AI can sift through large datasets, such as medical records and genetic information, to identify subtle patterns that flag someone’s risk for developing diabetes years in advance.

AI algorithms are being developed to identify specific metabolic subtypes of T2D, which can help doctors tailor treatment to someone’s underlying biology.

AI is also helping to detect and prevent diabetes complications before they became major issues.

  • Eyes: AI-powered systems are being used for diabetes retinopathy, a major complication affecting the eyes. These tools can analyze retinal images almost instantly to spot early signs of diabetes retinopathy with accuracy compared to eye doctors, and this speeds up screening and increases access to crucial eye care. Importantly, though, this is not a replacement for routine diabetes eye care.
  • Other complications: Researchers are using AI to analyze retinal photos and other data to predict the risk of other severe complications, including heart disease, kidney disease, or foot problems, allowing for earlier preventive care.

AI is becoming a powerful tool to help some people manage certain diabetes tasks. This can include medication monitoring and reminders, scheduling prescription refills or doctor appointments, or establishing an exercise plan or meal plan with recipes and ingredient lists. It may also help predict certain blood sugar changes or how medications or insulin may affect your blood sugar levels.

Despite the possible benefits, AI is not a replacement for healthcare professionals. Always consult your diabetes care team before making any changes to your diabetes management.