What changed since early 2025
For the first time, AI-enabled startups captured a majority (≈62%) of digital health VC in H1 2025, a clear signal that automation and analytics are moving from pilots to production. Rock Health Healthcare
At the same time, governance is crystallising. Under the EU AI Act’s phased timeline, obligations for general-purpose AI (GPAI) models became applicable on 2 August 2025, with broader high-risk system requirements following through 2026–2027. This raises the bar for transparency and oversight when AI is used in evidence generation, submissions, or payer engagement. Digital Strategy
Payers are deploying AI — and that changes access dynamics
Large US insurers have moved beyond pilots. Cigna launched a generative-AI member assistant to answer benefits/claims/care questions, escalating to humans when needed. Expect similar assistants across payer portals and call centres. Cigna Healthcare Newsroom
In June, 50+ plans pledged to streamline prior authorisation, including real-time decisions for at least 80% of electronic PA requests by 2027—a material shift for pull-through and access planning. ahip.org
Pricing & revenue operations: AI moves into the core toolset
On the manufacturer side, revenue teams are formalising AI in pricing and contracting workflows. In Model N’s 2025 State of Revenue report, 62% of life-sciences leaders say they use or plan to use generative AI for revenue management, aligning with what we see in the field: policy-aware scenario modelling in hours (not weeks) and tighter integration of RWD into price/discount strategy. Model N
Clinical AI maturity raises the evidence bar
Clinically, machine-learning tools are now mainstream: the FDA notes 1,000+ AI/ML-enabled medical devices authorised across specialties (radiology, cardiology, etc.). For PMA, this matters because accuracy and workflow gains feed directly into value and budget impact arguments. U.S. Food and Drug Administration
Policy & reimbursement signals to watch
In the US, CMS’s CY2026 Physician Fee Schedule (PFS) proposes significant updates to practice-expense methodology (shifting indirect PE between facility/non-facility) and seeks input on payment for digital therapy devices (e.g., ADHD), indicating how software-enabled care is being brought into core payment policy. Centers for Medicare & Medicaid Services
In the EU, GPAI obligations now apply; companies using LLM-powered tools in EU market access should map use-cases, documentation, and risk controls against the Act. Digital Strategy
What PMA leaders should do now
- Embed AI into pricing workflows. Operationalise multi-market, policy-aware scenario modelling and link outputs to payer decision points
- Design for payer AI. Assume dossiers and communications will be parsed algorithmically; prioritise clarity, structured logic and explainability
- Harden EU AI Act compliance. Catalogue AI use across the access lifecycle; prepare GPAI documentation for EU markets
- Monitor reimbursement mechanics. Track CMS and national frameworks as they evolve PE, digital therapy coverage and related policies