What happens if Anthropic and OpenAI start to compete directly in healthcare?

The Scale of the Threat

Both companies are now operating at a scale that makes them credible vertical competitors, not just infrastructure providers. Anthropic is approaching a $900B valuation ahead of its anticipated IPO, while OpenAI closed $110B at an $840B post-money valuation earlier this year. Both have formally declared healthcare a strategic product category, not a side effect of API usage.

 How They're Moving Into Healthcare

OpenAI is executing a dual-flanking strategy:

Launched ChatGPT Health (consumer) and OpenAI for Healthcare (enterprise) in January 2026, including HIPAA-compliant tools for clinical documentation, care pathway management, and referral letters

Launched ChatGPT for Clinicians, free for any verified NPI-registered physician, directly undercutting paid point solutions

Acquired Torch, a medical records startup, for ~$100M — analysts expect more acquisitions to follow
Hired Nate Gross (Doximity co-founder) as healthcare strategy lead and Ashley Alexander (Instagram VP) as VP of Health Products

Anthropic is going enterprise-first via operations and compliance:

Claude for Healthcare includes native connectors to CMS Coverage Database, ICD-10, NPI, PubMed, and FHIR — targeting prior authorization, billing, and revenue cycle workflows

Positioned as the "universal connector" across the entire healthcare ecosystem (providers, payers, patients)
Already embedded with Veeva, Commure, Viz.ai and major consulting partners (Deloitte, PwC, KPMG)
 
The Real M&A and Competitive Threat Landscape

Vertical Integration is Underway

OpenAI has explicitly signalled it wants to own not just foundation models but the application stack across verticals, including healthcare. The Torch acquisition signals that health data assets are now acquisition targets, not just partnership opportunities.

API Access Can Be Revoked and Has Been

The Brookings Institution documented concrete precedents:

Anthropic cut off a coding startup's access when it was about to be acquired by OpenAI (April 2025)

Anthropic cut off OpenAI itself from Claude when it suspected model benchmarking

Anthropic blocked xAI's access entirely

Anthropic's ToS explicitly forbids using its models to build apps that compete with Anthropic's own offerings

Brookings concluded bluntly: "This should be a wake-up call to developers that Anthropic, OpenAI, or Google can just cut off access to the AI model they need to have a viable AI application."

Market Concentration Creates Structural Risk

OpenAI, Anthropic, and Google together controlled ~90% of the $37B enterprise AI market at end of 2025. This oligopoly means there is limited redundancy for most healthtech companies — switching is technically and commercially expensive.

 Who is Most Exposed vs. Protected

High-risk healthtech categories (thin moats, easily replicated):

Ambient scribing / clinical documentation (DAX, Abridge, Ambience compete directly with OpenAI for Clinicians)

General health chatbots and symptom checkers

Prior authorisation automation

Medical records summarisation

Lower-risk / defensible categories:

Companies with proprietary longitudinal patient datasets that no foundation model can replicate

Deep EHR integrations with Epic, Oracle Health — switching costs are enormous

FDA-cleared or CE-marked diagnostic AI (regulatory moat is genuine)

Disease-specific models trained on scarce, proprietary imaging or genomic data

Companies embedded as workflow infrastructure (not just AI features on top)

As Breyer Capital noted, "startups selling into healthcare enterprises no longer compete only with the EHR vendor's roadmap — they also compete with what a foundation model provider can plausibly ship for that customer, often faster and at lower marginal cost."

To discuss how Nelson Advisors can help your HealthTech, MedTech, Health AI or Digital Health company, please email [email protected]