Can a Phone App Check a Mole? AI Skin Cancer Apps Explained

You notice a mole that looks a little darker than you remember. It’s late, the dermatologist’s next opening is weeks away, and an app store ad promises an AI risk check in 30 seconds. It’s tempting.

Here is the honest answer: a phone app can help you photograph, track and organize your skin checks, and some apps try to estimate risk. But no app should be the reason you decide a spot is safe. The evidence behind consumer skin cancer apps is thin, regulators have pushed back on overconfident claims, and missing a melanoma is a far bigger problem than an extra doctor’s visit.

This guide explains what these apps can and can’t do, what the research and regulators say, and how to use them safely. This is general information, not medical advice.

What skin-check apps actually do

Most skin cancer apps fall into 3 groups, and they work very differently.

The AI risk-rating apps get the most attention. They use image recognition models, usually trained on large sets of labeled skin photos, to look for features linked to melanoma and other skin cancers. In theory, that’s similar to what a dermatologist does when looking at a lesion. In practice, a phone photo in your bathroom lighting is a much harder input than a dermatoscope image taken in a clinic.

What the research says about accuracy

The best-known review is a 2020 systematic review in The BMJ by Freeman and colleagues. It looked at 9 studies covering 6 algorithm-based smartphone apps that assess skin cancer risk.

The findings were sobering:

• Reported sensitivity and specificity varied widely between studies.
• The studies were small and of poor methodological quality, with selective recruitment and high rates of photos that couldn’t be evaluated.
• In most studies, clinicians, not app users, chose the lesions and took the photos. The authors expected real-world performance by everyday users to be worse.
• They concluded the apps can’t be relied on to detect all cases of skin cancer, and that the European CE marking process for these apps did not adequately protect the public.

AI for skin images has improved since then, and research continues. But a newer model in a lab study doesn’t automatically mean a consumer app works well on your phone, your skin tone and your lighting. Researchers have also raised concerns that many skin image datasets include relatively few images of darker skin, which can make AI less reliable for some people. Our explainer on AI bias covers why training data matters.

Key takeaway: A “low risk” result from an app is not a clean bill of health. Treat apps as a reminder system and a photo diary. If a spot worries you, the app’s score shouldn’t change your decision to get it checked.

What regulators have said

The FTC and overconfident claims

In February 2015, the Federal Trade Commission settled with marketers of 2 apps, MelApp and Mole Detective, that claimed to assess melanoma risk from smartphone photos. The FTC said the companies lacked adequate scientific evidence for their accuracy claims. The settlements barred them from claiming an app can detect or diagnose melanoma unless the claim is backed by competent and reliable scientific evidence.

That case still shapes the market. Many apps now carefully say they are not diagnostic tools and are for “information” or “tracking” only.

The FDA and medical devices

Software intended to diagnose disease can be regulated by the FDA as a medical device. In our research, we found no consumer smartphone app authorized by the FDA to detect skin cancer. Some apps that are available in Europe with CE marking are not offered for diagnosis in the US.

The FDA has authorized AI-assisted tools for clinicians. In January 2024, it authorized DermaSensor, a handheld device that uses light and an AI algorithm to help primary care physicians evaluate skin lesions for the most common skin cancers. That’s an important distinction: the AI supports a trained clinician during an exam rather than giving consumers a verdict at home. For more on how AI is used in other areas of medical imaging, see AI in radiology: what patients should know.

Why apps can’t replace a dermatologist

A dermatologist does a lot that a phone photo can’t capture:

• Full-body context. They compare a spot with your other moles. A spot that looks different from all the others can be a warning sign even if it looks ordinary on its own.
• Dermoscopy. A dermatoscope magnifies and lights the skin to show structures invisible in a normal photo.
• Touch and history. Texture, tenderness, how fast it changed, sun exposure and family history all matter.
• Biopsy. Only a biopsy examined under a microscope can confirm skin cancer.
• Accountability. A clinician follows up, documents findings and acts on results.

Apps can also miss things they were never designed to see, like spots on your scalp, back, or under nails, or lesions that don’t fit the typical pattern of the cancers the model was trained on.

How to check your skin: the ABCDEs

The American Academy of Dermatology (AAD) recommends watching for the ABCDEs of melanoma:

• A is for asymmetry: one half doesn’t match the other.
• B is for border: edges are irregular, scalloped or poorly defined.
• C is for color: several shades, such as tan, brown or black, or patches of white, red or blue.
• D is for diameter: melanomas are usually larger than about 6 mm, roughly a pencil eraser, though they can be smaller.
• E is for evolving: the spot looks different from others or is changing in size, shape or color.

The AAD also advises seeing a board-certified dermatologist if you notice new spots, spots that look different from your others, or spots that change, itch or bleed.

Where to start: using a skin app the safe way

If you want to use technology for skin checks, here is a practical approach:

1. Choose a tracking-first app. Prioritize apps that make it easy to take consistent, dated photos and compare them side by side.
2. Read the claims carefully. Be wary of apps that promise to “detect” or “diagnose” cancer. Look for clear disclaimers and published clinical evidence.
3. Check the privacy policy. Skin photos can be sensitive, especially of covered areas. See whether photos are stored in the cloud, used to train AI or shared. Our guide on whether your health app is covered by HIPAA explains why many consumer apps fall outside HIPAA.
4. Take good photos. Use natural light, the same distance each time, and a ruler or coin for scale.
5. Do monthly self-exams. Use a full-length mirror and a hand mirror, and ask a partner to check your back and scalp.
6. Act on changes, not scores. Book an appointment for any new, changing, itching or bleeding spot, whatever the app says.
7. Consider teledermatology if access is tight. A dermatologist reviewing your photos remotely is a far better option than an algorithm alone. See telehealth and remote patient monitoring for how virtual visits work.

The bigger picture: AI as a helper, not a judge

Skin cancer apps are a good example of how AI in health care works best: as a helper inside a system with clinicians, not as a standalone judge. AI tools that support trained professionals are gaining regulatory approval and real-world use. Consumer apps that hand out verdicts have a weaker track record. Read more in our overview of AI in healthcare in 2026, or our guide on whether to ask an AI chatbot about symptoms.

Conclusion

A phone app can be a useful tool for photographing your moles and noticing changes over time. It can’t diagnose skin cancer, and a reassuring score should never talk you out of seeing a dermatologist. Learn the ABCDEs, check your skin monthly, track changes with photos, and get any new or changing spot examined by a professional.

Sources

• The BMJ: Algorithm based smartphone apps to assess risk of skin cancer in adults (Freeman et al., 2020) — systematic review of app accuracy and study quality
• FTC: Marketers of “melanoma detection” apps settlement (2015) — FTC action on MelApp and Mole Detective claims
• American Academy of Dermatology: What to look for, the ABCDEs of melanoma — warning signs and when to see a dermatologist
• npj Digital Medicine: Learnings from the first AI-enabled skin cancer device for primary care authorized by FDA — DermaSensor authorization in January 2024 for primary care use