SAN FRANCISCO – Days after OpenAI entered the medical artificial intelligence (AI) market, Anthropic on Sunday unveiled Claude for Healthcare at the J.P. Morgan Healthcare Conference here.
Anthropic’s bid to take on OpenAI’s ChatGPT Health underscores intensifying competition to capture the lucrative healthcare sector, in which enterprise contracts reach millions of dollars annually.
Action in the sector is coming fast and furious. On Monday, OpenAI Group PBC said it acquired Torch Health Inc., a healthcare technology startup specializing in medical data aggregation.
The financial terms of the deal remain unclear, with conflicting reports on the purchase price. The Information reported the acquisition cost approximately $100 million, while CNBC cited sources placing the figure at roughly $60 million. OpenAI did not officially disclose the transaction value.
Torch develops an AI-powered health application that consolidates fragmented medical information from disparate sources including healthcare providers, diagnostic laboratories and personal health records. The platform aggregates lab results, prescriptions, diagnoses and treatment histories into a unified, standardized interface aimed at addressing the record fragmentation that often leaves patients and clinicians operating with incomplete or outdated information.
Conversely, Anthropic’s new service covers consumer health record integration for individual subscribers and enterprise tools for healthcare organizations. Pro and Max subscribers in the U.S. can now connect their medical records through HealthEx, a startup that aggregates data from more than 50,000 health systems nationwide.
The consumer features use Anthropic’s Model Context Protocol (MCP) to retrieve specific portions of medical histories rather than entire records. When users ask health-related questions, Claude requests only relevant categories such as medications, allergies, or recent lab results. Apple Health and Android Health Connect integrations began rolling out in beta this week through Claude’s mobile applications.
“When navigating through health systems and health situations, you often have this feeling that you’re sort of alone,” said Eric Kauderer-Abrams, Anthropic’s head of life sciences. He described a vision where Claude orchestrates personal information, medical records, and insurance data to simplify healthcare navigation.
For healthcare providers, Claude for Healthcare addresses administrative burdens that consume clinical resources. The platform now includes connectors to industry databases including the Centers for Medicare & Medicaid Services Coverage Database, the National Provider Identifier Registry, and PubMed. Early adopters include Banner Health, Stanford Healthcare, and pharmaceutical companies Novo Nordisk, Sanofi, and AbbVie.
Enterprise applications focus on automating time-consuming tasks like prior authorization requests and insurance appeals. Dhruv Parthasarathy, chief technology officer at medical documentation company Commure, said the features will help save clinicians millions of hours annually.
Microsoft Corp. said Claude in Microsoft Foundry now includes healthcare-specific capabilities for enterprise customers already using its cloud infrastructure. The technical foundation is Claude Opus 4.5, which Anthropic claims performs better on medical tasks with fewer errors than previous versions.
Both Anthropic and OpenAI emphasize strict limitations on their healthcare tools. Anthropic’s acceptable use policy requires qualified professionals to review content before finalization when Claude is used for healthcare decisions, diagnosis, or patient care.
The competing services share similar privacy protections. Health data shared with Claude is excluded from the model’s memory and not used for training future AI systems. Users can disconnect or modify permissions at any time.
Dr. Nasim Afsar, a physician and former chief health officer at Oracle Corp., said ChatGPT Health represents the first layer of what she calls “intelligent health.”
“AI can now explain data and prepare patients for visits,” she said in an email. “That’s meaningful progress. But transformation happens when intelligence drives prevention, coordinated action, and measurable health outcomes, not just better answers inside a broken system.”
The launches come amid heightened scrutiny of AI chatbots in healthcare. Last week, Character.AI and Google settled a lawsuit alleging their AI tools contributed to worsening mental health among teenagers who died by suicide. Anthropic and OpenAI caution that their systems can make mistakes and should not substitute for professional medical judgment.
