January 14, 2026 Anthropic is pushing deeper into healthcare with a new suite of AI tools aimed at doctors, insurers and patients. The expansion, called Claude for Healthcare, is designed to tackle one of US medicine’s most persistent problems: administrative overload. Anthropic says the tools can reduce paperwork, speed up insurance approvals and help both clinicians and patients better understand complex medical information.
At the core of Claude for Healthcare is Anthropic’s latest flagship model, Claude Opus 4.5. The company says the model performs significantly better than earlier versions on simulated medical and scientific tasks while producing fewer factual errors.
Unlike generic AI assistants, Claude for Healthcare connects directly to established US medical infrastructure. Anthropic says the system can query databases including the Centers for Medicare & Medicaid Services Coverage Database, ICD-10 medical coding standards, the National Provider Identifier Registry and PubMed’s biomedical research library. The idea is to surface verified information quickly and support workflows that routinely slow down care.
One major target is prior authorization, the insurance approval process that often delays treatment. According to a 2024 American Medical Association survey, prior authorizations consume an average of 13 hours of physician and staff time per week.
“These tools can be used to speed up prior authorization requests so that patients can get life-saving care more quickly … and help with regulatory submissions so that more life saving drugs can come to market faster,” the company said.
The launch comes after OpenAI unveiled its own medical offering. There has been a surge of investor interest in healthcare AI, with startups such as Abridge and Sword Health reaching multibillion-dollar valuations and large model developers racing to move beyond general-purpose chatbots into domain-specific systems.
On the consumer side, Anthropic is allowing US users on its Pro and Max plans to connect Claude to personal health records. New integrations with HealthEx and Function Health launched in beta on Sunday, with Apple HealthKit and Android Health Connect rolling out this week through Claude’s mobile apps. Anthropic said health data accessed through these connectors is not stored in Claude’s memory or used to train its models.
For life sciences customers, Anthropic is expanding Claude’s reach into clinical research with connectors to platforms such as Medidata, ClinicalTrials.gov and bioRxiv. New agent skills are designed to help draft FDA- and NIH-compliant trial protocols and monitor study performance.
Whether healthcare truly needs AI beyond narrow administrative gains remains an open question. Previous attempts to deploy AI in diagnostics and imaging have struggled in real-world settings, often running into bias, liability concerns and integration failures.
