Healthcare CFOs Warn AI Investments Will Fail Without Data Overhaul

Healthcare's AI infrastructure problem

Healthcare organizations are investing heavily in artificial intelligence, but a fundamental mismatch threatens to derail those efforts. According to senior finance executives speaking at the Healthcare Financial Management Association's annual conference, the industry's decades-old technology infrastructure cannot support the sophisticated AI tools being deployed.

Seema Verma, former CMS administrator and now general manager of Oracle Health & Life Sciences, framed the challenge directly: organizations cannot execute an AI strategy without first establishing a coherent data strategy. The infrastructure most providers operate today lacks the capability to deliver what AI requires — unified, real-time data spanning clinical, financial, and operational systems.

Consider a physician using AI to recommend treatment. That model needs instant access to insurance coverage details, pharmacy inventory, and potential drug interactions. Without integrated systems providing that connective tissue, even powerful AI models will produce suboptimal results, Verma explained.

Why it matters

Healthcare organizations are pouring billions into AI while their underlying systems remain fragmented and outdated. This disconnect means many high-profile investments may yield disappointing returns, wasting capital that could address the data infrastructure problems blocking real transformation. The warning from CFOs managing some of America's largest health systems signals that the industry's AI gold rush risks becoming an expensive detour unless foundational work happens first.

Scale of technical debt blocks progress

Mike Marks, CFO of HCA Healthcare, one of the nation's largest health systems, emphasized that legacy systems represent an active barrier to transformation. The cost of replacing technology at enterprise scale is enormous, and demand for AI innovation already exceeds what health systems can afford to spend.

Marks outlined a prioritization framework: clinical systems first, followed by operations and administrative functions. His reasoning centers on patient impact. Clinical systems directly affect care delivery, so improvements there produce the most immediate health outcome benefits. Operational and administrative AI matters, but should optimize systems already delivering high-quality care rather than compensating for clinical deficiencies. Clinical AI failures carry direct patient consequences, demanding the most rigorous investment and oversight.

The bot-versus-bot stalemate

Scott Hawig, CFO of BJC Healthcare, described the current state of healthcare AI through the lens of payer-provider dynamics. His characterization: two sets of bots locked in endless combat over claims, with neither side gaining ground. Provider revenue cycle bots battle insurance denial bots in what Hawig called "the fundamental problem" — sophisticated technology deployed on both sides of a broken process, automating dysfunction rather than resolving it.

The assessment from these finance leaders is stark. Investment capital is abundant, AI tools exist, and urgency is real. But deploying advanced models atop fragmented, outdated infrastructure may simply be a more expensive way of remaining stuck in place.

These details were first reported by MedCity News, which covered the HFMA conference discussions.