AI-Driven Care Denials Put Spotlight on Algorithmic Accountability in Elder Health

When artificial intelligence enters the care authorization pipeline, the stakes for older adults become especially high. A new report from Think Global Health examines AI-driven pilot programs that have demonstrably delayed access to care for Americans, adding a critical counterweight to the prevailing industry optimism around AI adoption in senior care settings. For agetech professionals who have spent years advocating for smarter, faster care coordination, the findings serve as a sobering reminder that deployment speed and accountability cannot be decoupled.

The Technology

The programs in question use algorithmic decision-making to evaluate care eligibility and authorize services — functions that have historically been managed by human case managers and clinical reviewers. On paper, the value proposition is straightforward: AI can process large volumes of claims data faster than human reviewers, theoretically reducing administrative bottlenecks and freeing clinical staff for direct care. In practice, however, the Think Global Health analysis found that these systems are introducing new forms of delay, with authorization decisions lagging rather than accelerating when AI tools are inserted into the workflow. The mechanisms driving these delays remain under scrutiny, but they likely include edge cases the models were not trained to handle, integration friction with legacy health IT infrastructure, and appeals processes that were not redesigned to match AI-generated outputs.

Why This Matters

Older adults and people with disabilities are disproportionately affected by care authorization delays. Unlike younger populations navigating a single acute episode, many seniors depend on continuous, coordinated services — home health visits, durable medical equipment, post-acute rehabilitation — where a lag of even a few days can trigger a cascade of adverse outcomes. The report arrives at a moment when Medicare Advantage plans and Medicaid managed care organizations are actively piloting AI tools for prior authorization, a trend that regulators have begun watching more closely. The Centers for Medicare and Medicaid Services has already signaled interest in guardrails around algorithmic decision-making in coverage determinations, and this new evidence will likely accelerate that policy conversation.

What's Next

For vendors building AI tools for care management and utilization review, the immediate implication is that explainability and auditability are no longer optional features — they are market requirements. Payers and provider networks that have moved quickly to deploy AI authorization tools without robust human-in-the-loop checkpoints may face both regulatory scrutiny and reputational risk. The agetech sector would benefit from developing shared standards around how AI recommendations in care settings are documented, contested, and corrected.

As AI embeds more deeply into the infrastructure of elder care, the industry's long-term credibility will depend on its willingness to measure outcomes rigorously and course-correct when the evidence demands it.