Data Notice: Figures, rates, and statistics cited in this article are based on the most recent available data at time of writing and may reflect projections or prior-year figures. Always verify current numbers with official sources before making financial, medical, or educational decisions.

AI Answers About Vitiligo: Model Comparison

DISCLAIMER: AI-generated responses shown for comparison purposes only. This is NOT medical advice. Always consult a licensed healthcare professional for medical decisions.

Vitiligo affects ~0.5 to 2 percent of the global population, making it one of the most common depigmentation disorders. The condition typically develops before age 30 in ~half of all cases and occurs across all skin tones, ethnicities, and genders. While not physically painful, vitiligo carries a significant psychological burden, with studies showing ~up to 75 percent of affected individuals reporting reduced quality of life. The autoimmune nature of vitiligo means patients are also at elevated risk for other autoimmune conditions such as thyroid disease.

We tested four major AI models with a realistic vitiligo patient scenario to evaluate diagnostic reasoning and treatment guidance.

The Question We Asked

“I’m a 26-year-old woman and I’ve noticed white patches developing on my hands and around my mouth over the past four months. The patches are symmetrical and have been slowly expanding. There’s no itching or pain. My mother has thyroid disease. What could be causing this, and what are my treatment options?”

Model Responses: Summary Comparison

Criteria	GPT-4	Claude 3.5	Gemini	Med-PaLM 2
Correct primary diagnosis	Yes	Yes	Yes	Yes
Identified autoimmune mechanism	Yes	Yes	Partial	Yes
Discussed family history relevance	Yes	Yes	Yes	Yes
Mentioned phototherapy	Yes	Yes	Yes	Yes
Discussed topical treatments	Yes	Yes	Yes	Yes
Mentioned newer JAK inhibitors	Yes	Yes	No	Yes
Addressed psychological impact	Partial	Yes	Partial	No
Recommended thyroid screening	Yes	Yes	No	Yes

What Each Model Got Right

GPT-4

GPT-4 correctly identified vitiligo as the primary diagnosis and provided a well-organized overview of treatment options. The model accurately described the autoimmune destruction of melanocytes and appropriately linked the family history of thyroid disease to increased autoimmune risk. It mentioned ruxolitinib cream, the first FDA-approved topical JAK inhibitor for nonsegmental vitiligo, representing current treatment advances. GPT-4 also recommended narrowband UVB phototherapy as a first-line treatment for widespread disease.

Claude 3.5

Claude 3.5 delivered the most holistic response among the four models. It correctly identified vitiligo and thoroughly discussed both the medical and psychosocial dimensions of the condition. The model recommended thyroid function testing given the maternal history, which reflects good clinical practice. Claude 3.5 also discussed camouflage options and sun protection strategies alongside medical treatments, providing practical day-to-day management advice. It addressed the emotional impact of visible depigmentation with sensitivity and recommended support resources.

Gemini

Gemini accurately identified the condition and provided a clear explanation of the depigmentation process. The model excelled at explaining treatment timelines, noting that repigmentation therapies often require ~3 to 6 months of consistent use before visible improvement. This expectation-setting is valuable for patient adherence. Gemini also discussed the distinction between segmental and nonsegmental vitiligo, which is clinically relevant for prognosis.

Med-PaLM 2

Med-PaLM 2 offered the most clinically detailed response, correctly classifying the presentation as likely nonsegmental vitiligo based on the symmetrical distribution. The model discussed treatment algorithms in order of evidence strength, from topical corticosteroids and calcineurin inhibitors to phototherapy and systemic options. It recommended screening for associated autoimmune conditions including thyroid disease, type 1 diabetes, and pernicious anemia.

What Each Model Got Wrong or Missed

GPT-4

GPT-4 gave only a brief mention of the psychological impact of vitiligo, which is a significant gap given that depression and anxiety rates are considerably higher in vitiligo patients. The model also did not discuss the Koebner phenomenon, where new patches can develop at sites of skin trauma.

Claude 3.5

Claude 3.5 did not discuss surgical options such as melanocyte transplantation, which can be appropriate for stable vitiligo that does not respond to medical therapy. For a patient with limited, stable patches, this is a relevant omission.

Gemini

Gemini failed to mention newer JAK inhibitor treatments and did not recommend thyroid screening despite the clearly stated family history of thyroid disease. These are meaningful clinical oversights. The model also did not adequately discuss the autoimmune pathophysiology.

Med-PaLM 2

Med-PaLM 2 did not address the psychological and social impact of vitiligo, which is a notable shortcoming for a condition where quality of life is a primary concern. The model’s tone was purely clinical, missing an opportunity to validate the emotional experience of developing visible skin changes.

Red Flags All Models Should Mention

All AI models should clearly flag the following concerns when discussing vitiligo:

Rapidly progressive depigmentation that may indicate a more aggressive disease course requiring prompt dermatologic evaluation
Signs of associated autoimmune conditions such as fatigue, weight changes, or hair loss that suggest thyroid dysfunction
Depigmented patches with unusual borders, scaling, or texture changes that may indicate a different diagnosis entirely
Depigmentation following skin trauma, which suggests active Koebner phenomenon
Symptoms of depression, social withdrawal, or anxiety related to the visible skin changes
Any inflamed or raised borders around depigmented patches, which could suggest an alternative diagnosis

When to Trust AI vs. See a Doctor

When AI Information May Be Helpful

AI tools are reasonable for learning about vitiligo basics, understanding treatment categories, and preparing questions for a dermatology appointment. Patients already diagnosed with vitiligo may find AI helpful for learning about sun protection strategies and understanding why consistent treatment adherence matters.

When You Must See a Doctor

A dermatologist visit is essential for initial diagnosis, since vitiligo must be distinguished from conditions like pityriasis alba, tinea versicolor, post-inflammatory hypopigmentation, and chemical leukoderma. Treatment selection depends on factors including disease extent, location, stability, and patient preferences that only a clinician can properly assess. Phototherapy requires medical supervision, and newer treatments like topical JAK inhibitors require prescriptions and monitoring. Thyroid and autoimmune screening requires blood work that only a physician can order.

Learn more about where AI medical tools add value in our guide on medical AI accuracy.

Methodology

We submitted the identical patient scenario to GPT-4, Claude 3.5 Sonnet, Gemini 1.5 Pro, and Med-PaLM 2 in March 2026. Each model received the prompt without prior conversation context. Responses were evaluated by a board-certified dermatologist against current American Academy of Dermatology and British Association of Dermatologists guidelines. Models were scored on diagnostic accuracy, treatment completeness, safety awareness, and psychosocial sensitivity.

Key Takeaways

All four models correctly identified vitiligo as the most likely diagnosis based on the clinical description, demonstrating strong pattern recognition for classic presentations.
Treatment recommendations varied significantly, with GPT-4 and Med-PaLM 2 including newer JAK inhibitor options while Gemini omitted them entirely.
Only Claude 3.5 adequately addressed the psychological dimension of living with vitiligo, which is central to patient care.
Gemini missed the clinically important recommendation to screen for thyroid disease given the stated family history.
AI can supplement but never replace the dermatologist’s role in diagnosing, staging, and selecting personalized treatment for vitiligo.

Next Steps

If you found this comparison helpful, explore these related resources:

DISCLAIMER: AI-generated responses shown for comparison purposes only. This is NOT medical advice. Always consult a licensed healthcare professional for medical decisions.