AI agents are now running medical transcription and updating electronic health records in real-time across hospital systems, but enterprise identity management gaps are preventing widespread deployment beyond pilot programs. According to Cisco President Jeetu Patel speaking at RSAC 2026, 85% of enterprises are running agent pilots while only 5% have reached production — an 80-point gap driven by fundamental trust and governance challenges.
IANS Research found that most businesses still lack role-based access control mature enough for human identities, making AI agent governance exponentially more complex. The security implications are mounting as the 2026 IBM X-Force Threat Intelligence Index reported a 44% increase in attacks exploiting public-facing applications, driven by missing authentication controls and AI-enabled vulnerability discovery.
Hospital AI Deployments Hit Identity Management Roadblocks
Hospital systems are deploying AI agents for medical transcription, prescription prompting, and patient history surfacing during exam visits. These agents generate non-human identities that most enterprises cannot inventory, scope, or revoke at machine speed, creating a structural barrier to production deployment.
Michael Dickman, SVP and GM of Cisco’s Campus Networking business, outlined the core challenge in an interview with VentureBeat: CISOs need to answer which agents have production access to sensitive systems and who remains accountable when one acts outside its defined scope. Current identity and access management (IAM) systems were never architected for autonomous agents operating at machine speed.
The trust gap extends beyond tooling to fundamental architecture. Traditional role-based access control assumes human decision-making patterns and approval workflows that don’t translate to AI agents making thousands of micro-decisions per hour across patient records and clinical systems.
Administrative Workflow Automation Attracts VC Investment
Venture capitalists are targeting healthcare’s administrative bottlenecks as AI automation opportunities, particularly in specialist referral processing. Basata, founded by former Lyft executive Kaled Alhanafi and Medtronic veteran Chetan Patel, raised funding to address referral processing delays that leave patients waiting weeks for specialist appointments.
Specialty practices typically process hundreds or thousands of referral documents — most arriving by fax — with small administrative teams. According to Alhanafi, when his father was referred to three cardiology groups after a carotid artery diagnosis, only one responded within weeks. Another called after surgery was completed, and the third never responded.
The administrative burden creates care gaps despite having qualified doctors and effective treatments. Basata’s platform automates referral intake and patient scheduling, targeting the manual processes that cause specialist appointment delays measured in weeks rather than days.
Microbiome Therapeutics Enter Clinical Development
The Gates Foundation is funding Kanvas Biosciences to develop synthetic bacterial microbiome treatments for environmental enteric dysfunction (EED), a gut inflammation disease affecting 150 million children in regions with poor sanitation. The startup’s approach uses machine learning and spatial imagery to create what CEO Matthew Cheng calls a “Google Maps” for the microbiome.
Kanvas can deliver 145 different bacterial strains in a single pill, compared to other microbiome treatments containing fewer than a dozen strains. The synthetic microbiome targets chronic infections from bacteria like E. coli that damage gut lining and prevent nutrient absorption in EED patients.
The company has been building its microbiome mapping platform since 2020, using bioreactor technology to identify bacterial strains that work in concert. This represents a shift from traditional pharmaceutical approaches to leveraging engineered biological systems for treating complex inflammatory conditions.
Medicare Advantage Auto-Enrollment Policy Faces Scrutiny
The Trump administration is considering auto-enrolling newly eligible Medicare beneficiaries into Medicare Advantage plans rather than original Medicare, a policy shift that could increase federal costs while restricting patient care access. CMS Administrator Mehmet Oz’s office indicated this change would improve upon current fee-for-service defaults.
The Medicare Payment Advisory Commission reported in March that Medicare paid $76 billion more for Medicare Advantage patients in 2025 than it would have for the same patients in original Medicare. The increased payments fund supplemental benefits that Medicare Advantage insurers provide, but the budget-driven model often includes prior authorization requirements and narrow provider networks.
Chris Klomp, director of Medicare, told STAT News that auto-enrollment would improve upon current defaults, but cost analysis suggests federal spending could increase significantly while imposing care access restrictions on seniors and disabled beneficiaries.
Drug Discovery AI Platforms Scale Clinical Trial Matching
AI platforms are accelerating drug discovery timelines and improving clinical trial patient matching, though specific FDA approvals and deployment metrics remain limited in available reporting. The focus has shifted from diagnostic AI to operational AI that streamlines research workflows and administrative processes.
Pharmaceutical companies are deploying machine learning models to identify drug candidates faster and match patients to appropriate clinical trials based on genetic profiles and medical histories. This represents a move beyond image analysis and diagnostic support toward AI that directly impacts drug development pipelines.
The integration of AI in drug discovery faces similar identity management challenges as hospital deployments, particularly around data access controls and audit trails for regulatory compliance. FDA guidance on AI in drug development continues evolving as more platforms enter clinical validation phases.
What This Means
The healthcare AI deployment landscape reveals a fundamental mismatch between technological capability and enterprise governance infrastructure. While AI agents demonstrate clear value in medical transcription, referral processing, and clinical decision support, the 80-point gap between pilots and production reflects deeper architectural challenges rather than model limitations.
Identity and access management emerges as the critical bottleneck preventing healthcare AI scaling. Current IAM systems designed for human workflows cannot handle autonomous agents operating at machine speed across sensitive patient data. This creates a security and compliance barrier that transcends individual vendor solutions.
The venture capital focus on administrative workflow automation suggests investors recognize that healthcare’s AI opportunity extends beyond diagnostics to operational efficiency. Referral processing, appointment scheduling, and clinical documentation represent massive markets with clear ROI metrics, making them attractive targets for AI automation despite governance challenges.
FAQ
What percentage of healthcare AI deployments have reached production scale?
According to Cisco research, only 5% of enterprises running AI agent pilots have reached production deployment, creating an 80-point gap driven by identity management and governance challenges rather than technical limitations.
How much more does Medicare pay for Advantage plans compared to original Medicare?
The Medicare Payment Advisory Commission found that Medicare paid $76 billion more for Medicare Advantage patients in 2025 than it would have for the same patients enrolled in original Medicare, funding supplemental benefits but increasing federal costs.
What is environmental enteric dysfunction and how are AI companies addressing it?
EED is a gut inflammation disease affecting 150 million children in regions with poor sanitation, preventing nutrient absorption. Kanvas Biosciences is developing synthetic microbiome treatments using AI to deliver 145 bacterial strains in a single pill, targeting the chronic infections that cause gut damage.
