EU AI Act Annex I Readiness Gate for Device Software Teams

A pre-send interrogation gate for regulatory and engineering leads at small medical-device software companies who must sign off Annex I / high-risk AI readiness claims to notified bodies, distributors, or internal launch committees before August 2026. The buyer manually enters each concrete readiness claim they are about to make — for example, that risk controls are implemented, data governance is documented, human oversight is defined, or post-market monitoring is operational. AE stress-tests each claim against a structured constraint set, forces explicit evidence links, flags premise-conclusion gaps and concession laundering, and keeps a portable miss-pattern ledger of which commitment types later failed under external review or launch slippage. This is not generic compliance software: the wedge is the compressed, binary deadline created by the trilogue collapse, combined with AE's ability to run adversarial debate over concrete commitments and grade them later against real external resolution events. The product removes the recurring pain of sending overconfident readiness statements that trigger remediation cycles, launch delays, or notified-body pushback.

AE has a credible wedge as the final readiness-claim gate for small medical-device software teams because it reduces overconfident Annex I / high-risk AI assertions at the exact moment they create delay, remediation, and review risk.

Engine verdict: GATHER_MORE_SIGNAL (WORTH_SKIMMING). Real claim-review pain exists, but the AI-Act/2026 wedge is unproven; validate paid concierge demand before building product.