Q&A: SKiN Angels and AI's Growing Role in Dermatology, With Peter Lio, MD

Artificial intelligence (AI) is moving quickly from novelty to clinical tool in dermatology. A 2024 study comparing ChatGPT and physician responses to patient portal dermatology questions found the chatbot's answers rated favorably against physician-written replies for quality and empathy.¹ Comparative work in this area continues to find AI-generated responses performing on par with, and at times ahead of, dermatologist-written answers on accuracy and tone.

Early-stage innovation is advancing alongside this technology, increasingly channeled through physician-led investment networks. SKiN Angels, an angel-investor group made up of dermatologists and other clinicians in the skin space, has evaluated more than 300 early-stage companies and funded over 30 since its founding roughly two years ago.² The group frames its mission around shaping the field rather than financial return alone.

Clinicians face real tradeoffs as adoption accelerates. AI can ease documentation and messaging burden and support patient education, but it remains capable of catastrophic errors in health contexts, so human oversight stays essential for the foreseeable future. Overreliance also raises concerns about eroding the reading, writing, and clinical reasoning skills trainees need to build independently.

Peter Lio, MD, is clinical assistant professor of dermatology and pediatrics at Northwestern University Feinberg School of Medicine and founding director of the Chicago Integrative Eczema Center. He addressed both SKiN Angels and AI's clinical role during a session at LEADderm 2026. In the following Q&A interview, Lio spoke with HCPLive at LEADderm 2026 in Newport Beach, California, about both topics:

HCPLive: What is the title of the talk you're giving today at LEADderm?

Peter Lio, MD: Today we're going to be introducing SKiN Angels. It stands for the Strategic Knowledge and Investment Network Group, and we're a group of angel investor dermatologists and other people in the dermatology and skin space. Similar to the show Shark Tank, we have founders and companies at a very early stage come to us, and they're usually asking for advice, connections, and oftentimes early angel investments. We've been operating for about two years now, and it's been incredibly exciting. Well over 300 companies have presented to us, and we've funded more than 30 of them.

HCPLive: What does the presentation cover, and what are its major takeaways?

Peter Lio, MD: Today we're going to talk about the genesis of SKiN Angels: why it came about, why there's a need, and why so many people are getting interested in starting their own product lines, companies, and software services. All of these different people are coming to us, and it has been an incredibly exciting time, particularly in the skin space. Tech startups have their own ecosystem of venture capitalists, but if you're in the skin space, you can be floating around a little more, and this really includes everything from OTC products to drugs to even devices. We're going to present that, and a couple of our companies are going to give short pitches so people get a sense of what it feels like when they present an idea, and how it would feel to be in the shoes of someone who has to decide whether to support them, throw in with them, or pass.

HCPLive: What are your hopes for what SKiN Angels ultimately leads to?

Peter Lio, MD: The biggest thing about it, and we say this in our introductory video to members and potential members, is if you ask your financial advisor, they'd say this is crazy, don't do it. It's not about making money. That would be an amazing side effect if we have some big successes and can share that with the founders, but really, it's about shaping the world of dermatology, and ultimately shaping the world around us. We can take good people we believe in, that we trust, that we think are going to do something great, find their ideas at an early stage, and help fan the flames so they can do great things. Even if we can help just a little, sometimes that's all you need: a tiny bit of encouragement, a tiny bit of support, networking, and money can actually change the world. At least that's what I believe, and that's our central thesis.

HCPLive: Why did you choose to give a talk on AI in dermatology?

Peter Lio, MD: AI has been thrust upon me and all of us. Whether you like it or not, it's coming up a lot, and there are a lot of questions around it. For me, I'm always looking for solutions to problems and new tools, and wherever you stand on AI, there's no doubt it's a very powerful tool that can solve some of these problems. There are a couple of areas I'm looking at. One is patient support. I call this the guidance gap: what happens between doctor visits, where you see your patient, give them a plan, and then you might not see them again for weeks or months, and they have questions and concerns in the middle. What often happens is portal messages and phone calls, and that's really exhausting.

Being able to leverage AI in that capacity could be really powerful. The other piece is how we as doctors can use AI and prompting to learn, teach, and move ourselves forward. We think about the historical ways we've learned, from presentations at conferences, publications, and posters, but now we have new things, including prompting. Asking a good question of AI can be really important. If you ask a bad question, you might not get the wrong answer, but you might get a misleading one that's not quite what you wanted. So it's about refining this skill for doctors and people in medicine and healthcare in general to use AI as best we can.

HCPLive: What data points or key elements do you discuss in the talk?

Peter Lio, MD: Yes, I try to bring it to real life. We've done a couple of studies where we took real patient queries, cleaned them of any identifiable health information, and kept the essence of the question. We put those into different AI models, and then had an international group of dermatologists rank the responses. Spoiler alert: the AI is generally incredible. It does a great job, answers beautifully, in a balanced and careful way, and honestly, many times better than we ourselves as dermatologists would do. Like it or not, it's a powerful tool, and now we can show the data.

On the other hand, to keep it fair and balanced, I also discuss some of the catastrophic and dangerous errors AI can make in these contexts. When you're dealing with somebody's health, a mistake carries big ramifications, so we talk about some of the types of errors it can make. They're getting better, but even though it isn't human, it's still prone to errors, and there are specific kinds it can still make. We have to be careful, and ideally we're always going to need some human oversight, at least for the foreseeable future.

HCPLive: Are you concerned AI could outsource how students gain knowledge and experience?

Peter Lio, MD: I'm always worried, and the world is changing very rapidly. I don't know if we'll be able to predict the implications and impact on people and development, especially kids, but there's no doubt we can't escape it. We can't put the genie back in the bottle, but you're right, it can make people lazy. It can make people lose the ability to read carefully, because someone who's already established, who knows how to write and read carefully, can use it as a tool more naturally. They can correct it and say, that's goofy, I don't understand what you're saying there, fix this.

Whereas if you're just trusting it, or treating it as an infallible source, you're much more likely to make a huge catastrophic error and skip the groundwork yourself. We're going to need in-person, real-life sessions with people to really know what they know: here's a pencil and paper in a quiet room, I need to see what you actually know. Otherwise, it's so easy to just ask AI to tell you about an essay or help put it together, and that's a real danger we're going to have to face.

Editor's Note: This transcript has been edited for grammar and clarity using artificial intelligence tools. Lio had no relevant disclosures of note.

References

1. Reynolds K, Nadelman D, Durgin J, Ansah-Addo S, Cole D, Fayne R, Harrell J, Ratycz M, Runge M, Shepard-Hayes A, Wenzel D, Tejasvi T. Comparing the quality of ChatGPT- and physician-generated responses to patients' dermatology questions in the electronic medical record. Clin Exp Dermatol. 2024;49(7):715-718. doi:10.1093/ced/llad456.

2. Lio P. Introducing SKiN Angels: The Strategic Knowledge and Investment Network Group. Presented at: LEADderm 2026; August 29-30, 2026; Newport Beach, CA.