Healthcare’s AI Bubble: Inside the Gold Rush Where Everyone’s an “AI Company”

Walking the floor at HLTH 2025 felt like stepping into healthcare’s version of the dot-com boom—complete with inflatable unicorns, pickleball courts, and the faint scent of casino cigarettes mixing with AI promises. Beneath the Venetian Resort’s glittering surface, however, a more complex story unfolded: an industry simultaneously embracing artificial intelligence while grappling with its first signs of AI fatigue.

The Great AI Saturation

According to reports from the conference floor, the sameness became overwhelming. “Everyone is framing themselves as the most generic, enterprise-wide agentic AI solution. It makes me want to vomit,” one health system executive confessed anonymously. The sentiment echoed through the cavernous convention center, where nearly every booth touted some variation of “AI-powered platform for healthcare transformation.”

What’s particularly striking is how quickly healthcare has reached this saturation point. While the broader tech sector has been wrestling with AI bubble concerns for months, healthcare appeared somewhat insulated—until now. The data tells part of the story: Rock Health analysis indicates digital health startups raised $6.4 billion in VC dollars in the first half of 2025, with 62% flowing to AI-focused companies. But the conference atmosphere suggested we’ve reached peak AI messaging, if not peak AI capability.

The Incumbent Counterattack

Meanwhile, established players are making their moves in ways that should concern every healthcare startup founder. Medical records giant Epic, though physically absent from the show floor, cast a long shadow. The company announced in August it plans to sell its own AI tools, including an AI scribe that directly competes with Abridge—the same $5.3 billion startup that Epic previously invested in before divesting earlier this year.

Even more concerning for startups: the arrival of pure-play AI giants. OpenAI made its first significant HLTH appearance with new healthcare lead Nate Gross sharing the main stage with Microsoft, Google, and Nvidia. While Gross remained coy about specific plans, his mere presence signaled OpenAI’s serious healthcare ambitions. “In my portfolio, people view OpenAI and Anthropic much more as a threat than people have ever viewed Amazon or Microsoft,” NEA partner Blake Wu noted, according to conference reports. “They’ve just moved a lot faster.”

The Valuation Conundrum

The funding announcements at HLTH revealed both the optimism and potential excess in healthcare AI. Medical AI search startup OpenEvidence reportedly landed $200 million at a staggering $6 billion valuation—the conference’s largest disclosed round. Hyro, positioning itself as a “responsible AI agent platform,” secured $45 million led by Healthier Capital.

Some investors defended the spending spree vigorously. Lightspeed Venture Partners’ Dr. Brenton Fargnoli argued during a panel that he’s seeing healthcare startups “growing faster than you’ve ever seen before, more efficiently than any company ever before.” GV’s Frédérique Dame echoed the sentiment, suggesting that “as much capital as possible is very important to grow a business and get as much market share as possible.”

Yet this growth-at-all-costs philosophy feels increasingly risky as venture capital faces pressure across sectors. The healthcare AI market appears to be reaching a critical juncture where the easy money chasing AI buzzwords may soon dry up, leaving only companies with genuine clinical utility standing.

Beyond the Hype: Real Applications Emerge

Amid the AI fatigue, several bright spots suggested where genuine innovation might be occurring. Biotech and pharma companies had a notably stronger presence than in previous years, with AI leaders from GSK, Novartis, and City of Hope detailing concrete applications in research and clinical trial recruitment.

Women’s health also took center stage in multiple conversations, representing a welcome shift for a historically overlooked sector. Companies like Midi Health, featured in JPMorgan digital ads throughout the conference, are demonstrating that targeted solutions for specific healthcare needs may ultimately prove more valuable than generic “AI platforms.”

Perhaps most importantly, the conference revealed growing emphasis on responsible AI development. Spring Health debuted new benchmarking for mental health chatbots, while the American Heart Association announced an AI assessment lab in partnership with Dandelion Health. As Pamir Ventures’ Milad Alucozai observed, “Smart money is really thinking about startups that are worried about, how do we build this? It’s not about shipping the fastest product anymore, it’s about shipping responsibly.”

The Healthcare Specificity Challenge

History suggests that Big Tech’s healthcare ambitions often stumble on the industry’s unique complexities. As UPMC Enterprises’ Mary Beth Navarra-Sirio noted, “I’ve seen a lot of really great technologies fail from other industries because they don’t understand how it works in healthcare. You could have the best technology in the world, but if you really don’t understand how it needs to support a clinical process, you won’t win in this space.”

This represents both a challenge and opportunity for startups. While OpenAI’s Gross highlighted ChatGPT’s 800 million weekly active users—an unimaginable scale for any single healthcare company—the question remains whether pure technology companies can navigate healthcare’s regulatory complexity, reimbursement models, and clinical workflows.

What Comes After the Gold Rush

The HLTH conference ultimately revealed an industry at an inflection point. The initial AI gold rush is giving way to a more nuanced phase where differentiation, clinical validation, and responsible implementation will separate winners from also-rans.

The parallels to previous technology cycles are hard to ignore. Just as the dot-com boom separated Amazon from Pets.com, the healthcare AI boom will likely see companies with genuine clinical utility and sustainable business models survive while generic “AI solutions” fade away. The entrance of established players like Epic and potential disruption from AI giants only accelerates this shakeout.

What’s clear from HLTH 2025 is that healthcare’s AI journey is just beginning, even as the hype cycle may be peaking. The real work—building clinically validated, responsibly deployed AI that actually improves patient outcomes—remains ahead. And in an industry where mistakes can cost lives, that work can’t happen fast enough.