Healthcare AI Companies: A Structured Overview of the Industry Landscape

The term "healthcare AI company" now spans an unusually wide range of business models, product types, and regulatory situations. A company building FDA-cleared radiology triage software occupies a fundamentally different regulatory and commercial position than one selling ambient documentation tools to physician practices, or one using machine learning to accelerate preclinical drug discovery. Grouping them under a single label is convenient for market analysts but not especially useful for anyone trying to evaluate a specific product or understand a company's actual risk profile.

This page organizes the landscape by application domain, notes which segments have the most FDA-cleared products, and flags the structural differences that matter most for professional evaluation — regulatory status, evidence depth, deployment model, and funding stage.

How the Landscape Breaks Down by Application Domain

The FDA's public database of AI/ML-enabled medical devices shows that radiology accounts for the largest share of cleared AI products by a wide margin — the majority of the 900+ authorized AI devices listed as of mid-2026 are imaging-related. But that concentration doesn't mean radiology AI is the only segment worth tracking. Several other domains have grown substantially and carry distinct regulatory and evidence profiles.

Radiology AI Companies

Radiology is where the FDA-cleared AI device market is most developed. Companies in this segment range from large established players with broad portfolios — Aidoc, Nuance (Microsoft), Nanox, Intelerad — to more focused startups built around a single modality or finding type. The cleared device counts vary substantially: some companies hold a single 510(k) clearance for one indication; others have accumulated a dozen or more across chest X-ray, CT, MRI, and mammography.

What makes radiology AI companies structurally interesting from an evaluation standpoint is that the FDA clearance record is relatively accessible and the evidence base — while uneven — is more developed than in most other segments. Several radiology AI tools have been through prospective validation studies, and a smaller subset have published real-world deployment data from health system implementations.

That said, FDA clearance in radiology AI is heavily concentrated in the 510(k) pathway, which establishes substantial equivalence to a predicate device rather than requiring independent clinical trials. A cleared radiology AI product is not automatically a clinically validated one — those are separate questions that require looking at the evidence record separately from the authorization record.

Key Structural Characteristics

• Most radiology AI companies sell to health systems and radiology groups via direct contracts or through PACS/RIS integration partnerships.
• Revenue models vary between per-scan fees, site licenses, and bundled enterprise contracts — fee structure affects how deployment volume is reported and how ROI is calculated.
• Algorithmic bias concerns are documented in this segment: multiple studies have identified performance differences across patient demographics, particularly in chest X-ray and mammography AI.
• Post-market surveillance requirements under FDA's SaMD framework apply to cleared radiology AI products, but enforcement and public disclosure of post-market findings remain inconsistent.

Ambient Documentation and AI Scribe Companies

The ambient documentation segment — companies building tools that listen to clinical encounters and generate structured notes — has attracted significant investment and rapid health system adoption since 2023. Nuance DAX (Microsoft), Abridge, Suki, Ambience Healthcare, and DeepScribe are among the more prominent names, though the competitive field is crowded and changing quickly.

Most ambient AI scribe products are built on large language models and are not classified as FDA-regulated medical devices under current guidance — they're positioned as documentation assistance tools rather than clinical decision support. That distinction matters for how they're evaluated: there's no FDA clearance record to check, and the evidence base largely consists of implementation reports and vendor-commissioned studies rather than peer-reviewed clinical trials.

Adoption has been faster in this segment than almost anywhere else in healthcare AI — partly because the value proposition (reducing documentation burden) is immediately legible to clinicians, and partly because the procurement path doesn't require the same regulatory review as a cleared medical device. Health systems have signed enterprise contracts with ambient AI vendors without the evidence review process that would accompany a new diagnostic AI tool.

Clinical Decision Support Companies

Clinical decision support (CDS) AI covers a broad range: sepsis prediction models, early deterioration alerts, medication dosing recommendations, readmission risk scoring, and more. The regulatory situation here is more complicated than in radiology AI — whether a CDS tool is regulated as a medical device depends on whether it meets the criteria for "device-CDS" under FDA's 2019 guidance, which turns on whether it's intended to replace clinical judgment or merely inform it.

Companies like Sepsis Sentry, Dascena (now part of larger health IT portfolios), and Epic's internally developed predictive models all occupy this space. Epic is a particular case: it has deployed sepsis prediction and deterioration models across thousands of hospitals via its EHR, but it doesn't operate primarily as a healthcare AI company — its AI products are embedded features, not standalone devices.

The evidence quality in the CDS segment is uneven. Some tools — particularly sepsis prediction models — have been studied extensively, and the results are mixed. External validation studies have repeatedly shown that models trained on single-institution data degrade when deployed at different hospitals. Model drift over time is a documented issue that few vendors disclose proactively.

Pathology AI Companies

Computational pathology has seen a meaningful increase in FDA clearance activity since 2022. Paige.AI received the first FDA-authorized AI for prostate cancer detection on whole-slide images (De Novo authorization, 2021), and several other companies — including PathAI, Proscia, and Hologic's PathSight — have followed with cleared products or are in active review.

Pathology AI companies face a different deployment challenge than radiology AI: digital pathology adoption (scanning slides to whole-slide images) is still incomplete across U.S. hospitals, which creates an infrastructure dependency that limits the addressable market. Companies in this segment often need to sell or partner on the scanning hardware alongside the AI software.

Revenue Cycle and Administrative AI

Revenue cycle management (RCM) AI is one of the largest segments by deployed volume and investment, but it sits entirely outside the FDA device framework. Companies like Olive AI (now restructured), Cohere Health, Waystar, and dozens of smaller vendors apply machine learning to prior authorization, claims coding, denial management, and eligibility verification.

The lack of regulatory oversight in this segment doesn't mean the products are unimportant — prior authorization AI, in particular, has drawn significant scrutiny from CMS and Congress over concerns that automated denial systems may inappropriately restrict patient access to care. Several health insurers using AI-based prior auth tools have faced state-level investigations and class action litigation.

Drug Discovery AI Companies

Drug discovery AI operates on a different timeline and regulatory pathway than clinical AI. Companies like Recursion Pharmaceuticals, Insilico Medicine, Schrödinger, and Exscientia apply machine learning to target identification, molecular design, and preclinical screening. The AI itself isn't regulated as a medical device — the output (a drug candidate) eventually enters the IND and NDA pathway, but the AI platform doesn't.

Several AI-designed drug candidates have reached Phase I and Phase II clinical trials as of 2026, but none have yet completed Phase III and received FDA approval. The segment attracts significant venture investment — Recursion is publicly traded, Insilico has raised over $400M — but the commercial validation timeline is measured in years or decades, not quarters.

What to Check When Evaluating a Healthcare AI Company

The most common mistake in evaluating healthcare AI companies is treating FDA clearance as the primary quality signal. Clearance establishes that a device is substantially equivalent to a predicate (for 510(k)) or meets a new device type standard (for De Novo) — it doesn't establish that the tool improves patient outcomes in your specific clinical environment.

1. Regulatory status: Is the product FDA-cleared, exempt, or unregulated? Which pathway? What is the stated intended use in the authorization record?
2. Evidence base: Are there peer-reviewed validation studies? Were they prospective or retrospective? Was external validation performed on a population different from the training set?
3. Deployment record: Has the product been deployed at scale? Are there published implementation reports with outcome data — not just vendor case studies?
4. Algorithmic bias disclosures: Has the company published subgroup performance data across demographic groups? Are there known performance gaps by race, sex, age, or body habitus?
5. Funding and business stability: Is the company publicly traded, late-stage private, or early-stage? Health systems entering multi-year contracts with Series A companies face meaningful vendor stability risk.
6. Regulatory and legal history: Has the company received FDA warning letters, been involved in enforcement actions, or faced litigation related to its AI products?

Funding Stage as a Structural Signal

Funding stage matters for healthcare AI companies in ways that don't apply as directly in other software sectors. A pre-revenue Series A company selling AI diagnostic tools to hospitals is asking those hospitals to take on vendor stability risk alongside clinical risk. If the company runs out of runway before achieving sustainable revenue, the health system is left managing a product withdrawal from a clinical workflow.

The 2023–2024 period saw several high-profile healthcare AI company failures and restructurings — Olive AI, Babylon Health, and others — that left health system customers managing abrupt product discontinuations. Those cases have made procurement teams more attentive to financial stability, though the pressure to adopt AI quickly hasn't disappeared.

A Note on Market Size Estimates

Healthcare AI market size figures circulate widely — estimates for 2025 global market size ranged from $14 billion to over $45 billion depending on the source and methodology. These figures are not comparable to each other. Different analysts include or exclude drug discovery AI, administrative AI, and international markets differently. Some count platform revenue; others count only software licenses. Some include AI features embedded in EHR systems; others don't.

Using This Site to Research Specific Companies

The company profiles in this section are structured records, not articles. Each profile records the company's primary application area, FDA-cleared product count, funding stage, key clinical partnerships, and any notable regulatory or legal history. Profiles are updated when material changes occur — new clearances, acquisitions, funding rounds, or product withdrawals.

For any company with FDA-cleared products, the profile links to the relevant device records in the FDA AI Device Records section, where you can check the specific intended use, clearance pathway, and available real-world evidence status for each product. For companies with peer-reviewed clinical validation studies, cross-links to the relevant evidence appraisals are included in the profile.

The goal is to support the specific verification task — "what has this company actually cleared, what does the evidence show, and what is their regulatory history" — rather than to provide a narrative summary of why a company is interesting or promising.