Information Matters Report #IM108 — Sector Context Briefing. May 2026.
Download the full briefing — 29 pages —here
Twenty vendors are doing meaningful agentic AI work in healthcare as of May 2026. Unlike legal or coding — categories where the cohort can be read as a single competitive set — the healthcare cohort divides cleanly into four sub-segments, each with its own buyer, its own regulatory clock, and its own structural moat geography. Our new Sector Context Briefing profiles all twenty and maps the four-way split.
Headline findings
- Clinical AI scribes have reached scale, and the cohort top is consolidating. Abridge (~30% share, $773M+ raised), Ambience Healthcare ($1.25B valuation), and Suki AI (~10% share) account for over half the segment between them. The middle tier of the cohort faces the same structural compression we’ve seen in coding-agent and legal-AI middle tiers.
- Revenue-cycle and claims is where agentic AI delivers the clearest enterprise ROI today. Akasa (Black Book’s #1 most-promising healthcare RCM startup of 2025), Notable Health, and Adonis each offer human-in-the-loop architectures that automate prior-authorisation and denials workflows where the cost of an error is precisely measurable.
- AI drug discovery has consolidated post-Recursion-Exscientia merger. The combined entity at ~$1.8 billion combined valuation is the public reference point. Insilico Medicine’s Phase 2 IPF readout in 2026 is the cohort’s clearest binary on whether AI-led discovery converts capital into clinical evidence at scale.
- Payer-side prior-authorisation is the most regulated and most policy-driven segment. The CMS Interoperability and Prior Authorization Final Rule takes initial effect in 2026; Cohere Health has processed 15 million authorisations; CoverMyMeds is the legacy McKesson-owned incumbent. Regulatory clock plus payer-side procurement scale make this the segment with the longest buyer commitments.
- Foundation-provider entry into healthcare has begun but lags coding and legal. Microsoft Nuance DAX Copilot, Google Cloud Healthcare API, and Anthropic’s May 2026 clinical partnerships are the named entries. Healthcare’s regulatory perimeter slows the playbook that worked in coding and legal.
- The buyer-side picture is uneven. Health systems and large physician groups are well into procurement-grade evaluation; community providers are at pilot stage; payers are deploying at scale in prior auth and payment integrity but cautiously elsewhere.
Why the four-way split is the briefing’s central read
A vendor that is well-positioned in one healthcare sub-segment is often irrelevant to the others. The vendor a health system selects as its ambient-scribe partner is not the same vendor it selects as its RCM-automation partner, and neither is the vendor a payer standardises on for prior authorisation. The four sub-segments operate on different procurement cycles, different contract sizes, different risk thresholds, and different regulatory perimeters.
The regulatory clock matters disproportionately in healthcare. 2026 is a year of specific regulatory events: the CMS-0057-F initial provisions, the EU AI Act high-risk-system rollout in August, FDA framework evolution on AI/ML medical devices, and the first cohort of state-level AI-in-medicine legislation reaching enforcement. Each reshapes a specific sub-segment differently. The briefing maps the regulatory clock onto the cohort segment by segment.
For enterprise buyers committing to multi-year integration paths, the implication is sector-specific: a healthcare AI commitment is not a single decision but a sub-segment-specific decision, and the moat geography differs sub-segment by sub-segment.
Download the full briefing — 29 pages —here
Report #IM108. Martin De Saulles, Principal Analyst; Fiona, AI Research Analyst.
