Insurers, PBMs (the pharmacy benefit managers that negotiate drug prices for insurers), and hospital chains use AI in healthcare to extract revenue, consolidate markets, and control clinical labor. While sources like Rock Health (2025) and KFF Health News (2025) acknowledge AI’s impact, they underestimate how far AI is engineered to enforce cost-cutting, deskilling, and monopolization.
Mainstream discussions of AI in healthcare fixate on efficiency gains, workflow optimization, and regulatory oversight. They treat AI as an inevitable force to be managed rather than a deliberate mechanism of austerity. The dominant discourse—centered on bias, fairness, and misinformation—treats AI failures as technical flaws, when many are what insurers and hospital chains deploy the software to do.
The trajectory is a choice
The future of AI in healthcare is a political struggle over control, and calling it a technological certainty is part of that struggle. Healthcare IT News (2025) highlights AI’s potential benefits but doesn’t question the financial interests that dictate its deployment.
Recent studies in 2025 show that AI both enhances and restricts patient care: it improves clinician workflows, reduces administrative burdens, and increases patient engagement (Becker’s Hospital Review, 2025). But reports like those from the Deloitte Healthcare Report (2025) misrepresent cost-cutting measures as efficiency gains. They ignore how these so-called optimizations lead to service rationing, labor precarity, and harsher productivity demands.
AI-driven rationing was designed: it optimizes claim denials, intensifies labor exploitation, and entrenches vertical integration. Whatever “evolution” AI displays is set by the insurers, PBMs, and hospital executives who profit from restricting care and controlling clinical labor.
The problem is who controls AI, under what conditions, and for what purpose. National Nurses United (2025) correctly identifies AI’s role in cost-cutting and in optimizing hospital staffing. In financialized healthcare, run to produce returns for investors, AI is optimized to justify and enforce austerity. Patient outcomes are incidental.
Rock Health (2025) warns of AI consolidation but frames it as an inevitable industry trend rather than a calculated accumulation of power by entrenched corporate players.
Consolidating markets, disciplining clinicians
AI serves corporate consolidation as much as cost-cutting. Insurers and hospital chains use it to monopolize healthcare ecosystems and eliminate competition, and its cost-optimization models favor large, vertically integrated players, crushing independent providers. UnitedHealthcare used naviHealth’s nH Predict model to forecast post-acute stays. Optum bought naviHealth in 2020, so the insurer and the model’s maker both sit inside UnitedHealth Group. A federal lawsuit alleges the insurer overrode physicians and cut off nursing-home coverage on the model’s schedule, and a judge has let contract claims proceed.
AI also replaces clinical discretion with actuarial efficiency metrics. Physician scoring penalizes doctors for deviating from pre-approved, cost-contained treatment pathways. Surveillance and deskilling force clinicians into data-entry roles while their high-value tasks get automated, and workforce-optimization software adjusts hospital staffing in real time to prioritize financial efficiency over patient need.
Half the nurses surveyed by National Nurses United say their employer runs algorithmic systems over EHR (electronic health record) data to score patient acuity and set nursing hours. Sixty-nine percent of those nurses say their own assessments don’t match the software’s acuity score. Among nurses in facilities that use outcome-prediction scores, 40% say they can’t modify the scores to reflect their clinical judgment. HCA Healthcare runs nurse and support staffing about 30% below the national hospital average across its 182 hospitals.
Taking AI out of corporate control
AI within a financialized system can be made “more accountable” and still serve the system. It needs to be removed from corporate control entirely. Reports from 2025, such as those from the AMA and Stanford Medicine AI Report, acknowledge that AI-driven cost-cutting measures lead to more clinician burnout and patient dissatisfaction. But these discussions don’t recognize that AI structures those conditions as much as it aggravates them.
Removing it from corporate governance would mean:
- Public, clinician-led AI governance models that prioritize patient care over cost-cutting.
- Abolition of AI-driven utilization management (an insurer’s review of whether a treatment will be covered, before or after it’s given) and denial algorithms.
- Reinvestment in human-centered care instead of AI-driven austerity measures.
- Prohibiting AI-based labor surveillance that fuels clinician burnout and deskilling.
What’s at stake is who controls healthcare, financialized actors or the public. AI as deployed in healthcare today is algorithmic austerity: an infrastructure built to prioritize financial goals over human well-being; fairness tweaks leave that infrastructure standing. It makes rationing harder to contest, then calls that efficiency, and it guarantees profits for insurers, PBMs, and hospital chains while patients’ costs stay put.
The clinical decision-support layer is the clearest case: “money-mangled” CDSS. The systems clinicians are told to trust are trained on billing data and liability aversion before clinical reality—so the recommendation that looks technical is already an insurance ledger wearing an evidence coat. Nobody wants a decision-support tool whose silent premise is the payer’s margin. An oversight board or a transparency report can leave that tool in service, and a fairness audit can pass it with the payer’s margin still built in.