AI Is Coming for Pharmacy Benefit Managers: Potential Winners and Losers

The article argues that AI is compressing PBM economics, with Cigna’s Express Scripts/Pharmacy Benefit Services most exposed as rebate transparency and client churn pressure margins. CVS and UnitedHealth are using AI to defend workflows, while Waystar is positioned as the clearest AI-native beneficiary; Elevance has the smallest PBM exposure. Despite strong Q1 beats and raised FY26 guidance across several names, the piece highlights regulatory, reimbursement, and Justice Department risks that could shift relative performance.

Analysis

The real shift is not “AI helps healthcare” but that AI attacks the hidden toll booths inside PBM economics: rebate arbitrage, manual authorization friction, and opaque claims exceptions. That compresses the wedge between gross drug spend and net realized margin, so the first-order losers are the names most dependent on spread capture, but the second-order impact is broader: lower administrative friction should raise patient switching and reduce lock-in, which weakens incumbents’ ability to defend formulary share with service complexity. In that regime, scale becomes less of a moat and more of a legacy-cost burden unless it is paired with superior data integration.

The market is likely underestimating how uneven the transition is. Near-term, the pressure shows up first in client renegotiations and slower growth, not in a sudden earnings collapse, because PBM contracts reset over months to years and large customers move cautiously. The bigger risk is that transparency and AI-assisted adjudication make “good enough” PBM services commoditized, forcing margin compression into a multi-year glide path while the vendors building workflow software monetize the disintermediation on a faster adoption curve.

Among the names here, the most interesting setup is the asymmetry between operational strength and valuation. The diversified managed-care/PBM incumbents can offset some compression with scale, but they also face the highest probability that AI adoption simply passes savings through to clients rather than preserving spread for shareholders. Meanwhile, the AI-native beneficiary has cleaner upside if conversion rates remain high; the main concern is whether customer concentration and leverage limit the rerating even if fundamentals stay strong.

The contrarian read is that this may be more of a margin-reallocation story than a pure destruction story. If AI reduces authorization and claims leakage, payers with the best data pipes could actually expand their value share by owning the orchestration layer, while independent workflow vendors capture only a slice of the savings. That makes the key question not “who has AI?” but “who controls the decision point when the rebate and authorization workflow gets automated?”

Key Decisions for Investors

• Long WAY on weakness over the next 2-4 weeks; use the dislocation between fundamentals and price as the entry, targeting a 25-40% rerating if execution holds, but cap downside with a stop if client additions or retention decelerate materially.
• Pair trade: long WAY / short UNH or CVS into the next 1-2 quarters. The thesis is that workflow automation monetizes faster at the software layer than it preserves economics inside incumbent PBMs; expected relative outperformance should come from multiple expansion rather than earnings alone.
• For more defensive positioning, prefer ELV over UNH in managed care over the next 3-6 months. ELV has the smallest PBM exposure and therefore the least direct margin compression risk if AI-driven transparency accelerates rebate pass-through.
• Avoid initiating fresh longs in CI on PBM optionality until there is evidence that client attrition and rebate erosion are stabilizing. This is a ‘show me’ story with asymmetric downside if large accounts reprice in the next renewal cycle.
• Consider a call spread on CVS for 3-6 months rather than outright equity. CVS can defend better than pure PBM peers, but the stock likely needs visible operating leverage from AI deployment to justify further multiple expansion from here.