Inside the "Dreamforce of healthcare," where AI hype and fear were hand in hand (www.businessinsider.com)

At HLTH 2025 in Las Vegas, AI dominated every booth and panel — from flashy “agentic” voice and enterprise assistants to an entire AI Zone — even as attendees voiced fatigue and fear of an overheating market. Investors poured capital into healthcare AI (digital health startups raised $6.4B in H1 2025, with 62% going to AI firms), fueling big rounds such as OpenEvidence’s $200M at a $6B valuation and Hyro’s $45M. But founders and VCs are nervous: incumbents (Epic) and hyperscalers/new AI giants (OpenAI, Anthropic) are racing into health, threatening startup moats, while many vendors offer homogeneous, unproven “platform” pitches that frustrate health system buyers. For the AI/ML community the conference underscored clear technical and market implications. Large language models and agentic systems promise scale and longitudinal consumer data advantages (ChatGPT’s massive weekly user base), but bring high-stakes risks like hallucinations and integration challenges with clinical workflows and EHRs. Responses at HLTH signaled a maturation: domain-specific LLMs (Claude for Life Sciences), new validation efforts (AHA’s AI assessment lab, Spring Health’s LLM benchmark for mental‑health bots), and investor emphasis on responsible, clinically grounded deployments. The takeaway: technical rigor, validated metrics, and workflow-aware models will determine winners as capital and competition concentrate.