Aidoc's Generative AI Push Tests FDA's Medical Device Rules

A Radiology AI Company Moves Toward Uncharted Regulatory Territory

Aidoc, an Israeli medtech company known for AI tools that flag abnormalities in medical scans, is preparing to bring a generative AI system to the FDA — one capable of interpreting images and drafting preliminary reports for radiologists to review 1. The move is notable because it comes at a moment when regulators, hospitals, and AI developers are still working out how generative AI, which can produce novel text and content rather than simply classify data, should be evaluated for safety and efficacy in clinical settings 1. Aidoc CEO Elad Walach has framed safety as the central concern, acknowledging that generative tools introduce risks and uncertainties that traditional diagnostic algorithms do not 1.

Regulators Are Already Approving Narrower AI Tools

While generative AI remains a regulatory gray zone, the FDA has continued clearing more conventional, narrowly scoped AI diagnostic tools. iHealthScreen recently received FDA clearance for AI software that screens for diabetic retinopathy by analyzing images captured through the iCare DRSplus camera, a device already common in ophthalmology clinics and optician shops across the country 2. That clearance illustrates the current path most medical AI products take to market: task-specific tools trained to detect a single condition, evaluated against a defined and relatively narrow set of clinical outcomes.

Critics Say the Approval Process Itself Is Outdated

Not everyone believes that existing pathway is adequate even for these narrower tools, let alone for generative systems. An op-ed on FDA device review argues that current testing and approval methods for AI-based healthcare products lag behind the technology itself, undermining confidence in how well-vetted these tools really are before reaching clinicians and patients 4. This tension — approving AI fast enough to keep pace with innovation while ensuring rigorous validation — sits at the core of the debate Aidoc's application is expected to intensify.

Coverage and Reimbursement Rules Are Shifting Too

Regulatory clearance is only part of the equation; getting paid for AI tools is another. Reporting from STAT notes that proposed policy changes could make Medicare coverage easier for AI devices that earn the FDA's "breakthrough" designation, potentially granting them automatic coverage rather than requiring separate reimbursement reviews 5. Such a shift would materially affect how quickly hospitals adopt new AI diagnostic and reporting tools, including systems like Aidoc's.

Broader AI Momentum in Medicine

The healthcare AI landscape extends well beyond diagnostics and reporting. Separate coverage of AI's role in drug discovery describes ambitions to cut development timelines and costs by as much as 70%, reflecting a broader industry narrative that artificial intelligence is reshaping multiple corners of medicine simultaneously 3. Taken together, these developments suggest regulators face mounting pressure on several fronts at once: validating new categories of AI like generative reporting tools, modernizing approval standards, and adapting reimbursement policy — all while the underlying technology continues to advance faster than the rules governing it.