Artificial Intelligence Healthcare Companies: A Structured Overview of the Active Landscape

The healthcare AI company landscape is genuinely difficult to map. Hundreds of companies claim clinical AI products, but the meaningful subset — those with FDA-authorized devices, active clinical deployments, or peer-reviewed evidence — is considerably smaller. This overview focuses on that subset, organized by primary application domain rather than by market capitalization or media attention.

The organizing principle here is the same one used across this site: what has the company actually cleared, deployed, and published? Vendor positioning and press releases are not treated as evidence. Where FDA authorization records exist, they are referenced directly. Where peer-reviewed studies support product claims, they are noted with their design and limitations. Where neither exists, that absence is stated.

How This Landscape Is Organized

Companies in this overview are grouped by their primary application area. A company's primary area is determined by where the majority of its FDA-cleared devices, disclosed revenue, or published clinical evidence sits — not by how the company describes itself in marketing materials.

• Medical imaging AI (radiology, pathology, ophthalmology)
• Clinical workflow AI (ambient documentation, EHR-embedded decision support)
• Genomics and precision medicine
• Drug discovery and development
• Revenue cycle and administrative AI

Some companies span multiple domains. Where that is the case, the primary area reflects their largest cleared product portfolio or their most substantiated clinical evidence base.

Medical Imaging AI Companies

Imaging AI is the most densely regulated segment of healthcare AI. Roughly three-quarters of all FDA-cleared AI/ML-enabled medical devices fall under radiology, pathology, or related imaging specialties. This concentration reflects both the maturity of the underlying computer vision methods and the relatively well-defined clinical tasks — detecting a pulmonary nodule, flagging an intracranial hemorrhage — that lend themselves to prospective validation.

Aidoc

Aidoc focuses on AI-powered radiology triage and workflow prioritization. The company holds multiple FDA 510(k) clearances covering intracranial hemorrhage, pulmonary embolism, cervical spine fracture, and aortic pathology detection. Its platform integrates with PACS systems and is designed to surface urgent findings to radiologists before routine reads.

Published deployment data from hospital systems has shown measurable reductions in time-to-treatment for stroke and PE cases, though most of this evidence comes from retrospective or pre-post implementation studies rather than randomized controlled trials. Aidoc is privately held and has raised over $100 million in disclosed funding as of early 2026.

Viz.ai

Viz.ai holds FDA clearances for stroke, aortic disease, and pulmonary embolism detection, with a platform that routes alerts directly to on-call specialists via mobile notification. The company has published prospective evidence on door-to-treatment time improvements in large vessel occlusion stroke, including data from multicenter deployments.

Viz.ai's business model depends on demonstrating measurable clinical impact, and their published evidence base is more prospective than most imaging AI peers. That said, the majority of outcome studies involve surrogate endpoints (time-to-notification, door-to-puncture time) rather than mortality or functional outcome at 90 days.

Paige

Paige operates in computational pathology, with FDA De Novo authorization for Paige Prostate — the first AI-based software to receive De Novo authorization for pathology slide analysis. The tool is designed to assist pathologists in detecting prostate cancer on whole slide images.

The company has since expanded its portfolio to include breast, colorectal, and lung cancer detection tools, though the regulatory status of these additional products varies. Paige is privately held and has disclosed partnerships with major academic medical centers for model development and validation.

iCAD

iCAD is a publicly traded company (NASDAQ: ICAD) with a long history in mammography computer-aided detection. Its current products include AI-based density assessment and cancer risk scoring tools for mammography, with multiple FDA clearances. The company has published performance data on sensitivity and specificity improvements in screening mammography workflows.

iCAD's equity considerations are worth noting: mammography AI performance has been documented to vary by breast density and patient demographics in several independent studies, and the company's published data does not uniformly disaggregate performance by race or ethnicity.

Arterys (acquired by Tempus)

Arterys was an early entrant in cardiac and oncology imaging AI, holding FDA clearances for cardiac MRI quantification and liver lesion characterization. The company was acquired by Tempus AI in 2023. Its cleared products continue to operate under the Tempus umbrella, though the product roadmap post-acquisition has not been fully disclosed.

Clinical Workflow AI Companies

Workflow AI encompasses a heterogeneous category: ambient documentation tools that transcribe and structure clinical encounters, clinical decision support systems embedded in EHRs, and NLP-based tools that extract structured data from unstructured clinical text. The regulatory picture here is more complex than imaging AI — many workflow tools are not classified as medical devices and therefore do not require FDA clearance.

Nuance (Microsoft)

Nuance Communications, acquired by Microsoft in 2022 for approximately $19.7 billion, is the dominant player in ambient clinical documentation. Its DAX Copilot product (formerly Dragon Ambient eXperience) uses large language model technology to generate structured clinical notes from ambient encounter audio. The product is integrated with Epic, Cerner, and several other major EHR systems.

Published evidence on DAX includes vendor-disclosed time savings data (typically 5–7 minutes per encounter in disclosed studies) and physician satisfaction metrics from early adopter health systems. Independent peer-reviewed evaluation of note accuracy and clinical safety is limited as of mid-2026. DAX Copilot is not FDA-cleared.

Abridge

Abridge is a privately held ambient AI documentation company that has disclosed a major deployment partnership with UPMC and subsequent expansions to other health systems. Its technology generates structured clinical notes from encounter audio using generative AI methods.

Abridge has published some peer-reviewed evidence on documentation quality and physician experience, which distinguishes it from several competitors who rely primarily on vendor-disclosed metrics. The company raised a $150 million Series C round in early 2024. Like all ambient AI scribe tools, it is not FDA-cleared.

Suki

Suki offers an AI-powered voice assistant for clinical documentation, targeting ambulatory and specialty care settings. It integrates with Epic, Athenahealth, and several other EHR platforms. Published evidence is primarily vendor-disclosed; independent prospective studies are limited. The company is privately held and has disclosed Series C funding.

Cohere Health

Cohere Health focuses on AI-assisted prior authorization — a specific and high-friction administrative workflow. The company's platform uses NLP to review prior authorization requests against payer clinical criteria and route decisions. It is not an FDA-regulated device. Published evidence on authorization approval rates and turnaround time comes primarily from payer and health system partners.

Genomics and Precision Medicine AI

Genomics AI companies apply machine learning to variant interpretation, polygenic risk scoring, and treatment matching. The regulatory environment here is distinct from medical imaging: many genomic AI tools operate under laboratory-developed test frameworks or companion diagnostic pathways rather than standard SaMD clearance.

Tempus AI

Tempus AI (NASDAQ: TEM) went public in June 2024. The company operates a genomic data platform used in oncology for molecular profiling, treatment matching, and real-world evidence generation. It holds FDA clearances for several oncology-related AI products and acquired Arterys in 2023, adding medical imaging capabilities.

Tempus has published extensively on its genomic data assets and has disclosed partnerships with pharmaceutical companies for clinical trial matching and drug development. Its AI-based treatment matching tools are used in oncology practices, though independent validation of match quality and clinical outcome improvement is an ongoing area of research.

Illumina (DRAGEN)

Illumina's DRAGEN platform applies AI/ML methods to secondary genomic analysis — variant calling, structural variant detection, and RNA-seq analysis. DRAGEN is integrated into clinical sequencing workflows at major genomics laboratories. Illumina is publicly traded (NASDAQ: ILMN) and holds multiple FDA authorizations for sequencing-based diagnostics, though DRAGEN itself operates primarily as an analytical pipeline rather than a standalone cleared device.

Drug Discovery AI Companies

Drug discovery AI sits upstream of clinical deployment — the products are not patient-facing medical devices, and FDA clearance in the traditional SaMD sense does not apply. What matters here is pipeline progress: how many compounds identified or optimized by AI methods have entered clinical trials, and what are the Phase I/II results?

Recursion Pharmaceuticals

Recursion (NASDAQ: RXRX) uses high-throughput imaging and machine learning to identify drug candidates at scale. The company has disclosed a pipeline of AI-identified compounds in Phase I and II trials across rare diseases and oncology. Its approach generates large phenotypic datasets from cellular imaging, which are then used to predict compound activity.

Recursion has faced scrutiny over the translation rate from AI-identified candidates to clinical success — a challenge shared across the drug discovery AI sector. As of mid-2026, no Recursion compound has reached Phase III or received FDA approval, though the pipeline remains active.

Insilico Medicine

Insilico Medicine uses generative AI methods for target identification and molecule design. The company's lead compound ISM001-055, an AI-designed molecule for idiopathic pulmonary fibrosis, has advanced through Phase II trials — one of the first AI-designed molecules to reach this stage with published interim data. The company is privately held.

Large Health System and EHR Vendors with AI Products

A distinct segment of the healthcare AI landscape is EHR vendors and large health IT companies that have embedded AI into existing clinical platforms. These are not pure-play AI companies, but their AI products are deployed at scale and warrant separate treatment.

Epic Systems

Epic has integrated multiple AI models into its EHR platform, including sepsis prediction (the Epic Sepsis Model, or ESM), deterioration index, and readmission risk scoring. The ESM has been the subject of significant independent scrutiny: a widely cited prospective study published in JAMA Internal Medicine found the model's performance in external validation substantially lower than vendor-disclosed figures, with AUC of 0.63 in the prospective cohort versus 0.76–0.83 in Epic's internal validation.

This gap between internal and external validation is one of the most important documented examples of model performance degradation in real-world deployment. Epic's AI tools are not FDA-cleared — they operate as clinical decision support software under the CDS exemption framework.

Oracle Health (formerly Cerner)

Oracle Health has developed AI-based clinical decision support tools embedded in its EHR platform, including deterioration prediction and medication safety alerts. Following Oracle's acquisition of Cerner in 2022, the product roadmap has emphasized integration with Oracle's broader cloud infrastructure. Published independent evidence on Oracle Health's AI-specific tools is sparse compared to Epic.

Landscape Comparison: Selected Companies by Domain and Regulatory Status

What Distinguishes Companies Worth Tracking

Across this landscape, a few dimensions consistently separate companies with substantiated clinical products from those primarily operating on market positioning:

• FDA authorization status: A 510(k) clearance establishes substantial equivalence to a predicate device — it is not a clinical efficacy determination. But it does require disclosure of intended use, performance testing, and labeling. Companies with cleared devices have a documented regulatory record that can be traced.
• External validation: Performance metrics derived solely from internal training and validation datasets are not generalizable. Companies that have published prospective or multi-site external validation studies are meaningfully more credible than those relying on internal benchmarks.
• Equity and subgroup data: Published disaggregation of performance by race, ethnicity, sex, and age is rare but increasingly expected. Companies that report subgroup performance — even when results are unflattering — demonstrate a higher standard of transparency.
• Post-market surveillance: For cleared devices, FDA requires post-market reporting of malfunctions and adverse events through the MDR system. Companies with a clean MDR history and active post-market study programs are preferable to those with unresolved adverse event reports.

Funding Stage and What It Signals

Funding stage is a rough proxy for company maturity, not product quality. Several well-funded companies have thin evidence bases; some smaller Series A companies have published stronger clinical validation than their larger peers.

That said, funding stage does affect deployment trajectory. A Series A company with a single cleared product and 50 hospital customers has a different risk profile for a health system considering adoption than a publicly traded company with 1,000+ deployments and an established post-market surveillance program.

Acquisitions and Consolidation

The healthcare AI company landscape has seen significant consolidation since 2022. Notable acquisitions include Microsoft/Nuance (2022), Tempus/Arterys (2023), and several smaller imaging AI companies absorbed into radiology IT platforms. This consolidation has mixed implications for clinical users:

• Cleared devices typically retain their authorization post-acquisition, but the intended use and labeling may change if the acquirer files a new submission.
• Post-market surveillance obligations transfer to the acquirer, which may affect the continuity of safety monitoring programs.
• Product roadmaps often shift after acquisition, sometimes discontinuing products that health systems have deployed and integrated.
• Pricing and support models frequently change, which has operational implications for health systems with multi-year contracts.

Scope and Limitations of This Overview

The companies described here represent illustrative examples across major domains, not a comprehensive market map. Readers seeking procurement guidance should consult the FDA device registry and the clinical application briefs for task-specific evidence summaries rather than relying on company-level profiles alone.