← All 95 filings
Xiangyu Guo (Independent Researcher)
Public / patientsMember of the publicFiled September 13, 20265,660 words · 1 attachmentFDA-2026-N-7874-0092
What they argued
RecovryAI’s one-line reading of the filing.
M1 from his section 4 and the attached EON specification: generation, assigned authority and permission to execute should stay distinct, high-consequence transitions need explicit human authorization gates and traceable records, and outputs that directly authorize or materially condition consequential clinical action require explicit human sign-off. M5 from the same section: material changes in how a device receives information, exercises authority or affects clinical action - including a new foundation model, memory, data source, tool integration, agent capability or execution pathway - should trigger proportionate reevaluation. autonomy_low is inform from the attachment's low risk tier (general information, educational explanation, low-stakes summarization released immediately under sampling); autonomy_high is direct because critical outputs require named human sign-off before release or execution. He addresses clinician review sampling and record density but not premarket evidence proportionality, the premarket-postmarket trade, or competency benchmarking, so M2-M4 are N. Type: he self-describes as an independent researcher with no institutional affiliation and filed under the Individual Consumer category, so public is used rather than academic.
Themes it raises
7 of the 21 themes in the docket.
Watching the device after it shipsFDA Q19, Q20
Who is accountable when something goes wrongFDA Q21
Controlling a device that keeps changingFDA Q22, Q23, Q24, Q25
Devices that plan and take actionsFDA Q26
Whether human oversight is real oversightFDA Q3, Q4, Q14, Q20, Q21, Q26
Records that let investigators reconstruct an eventFDA Q19, Q21, Q24, Q26
How this fits rules that already existFDA Q8, Q9, Q16, Q25
FDA questions it names
Questions this filing names by number.
Q19 · Postmarket performance evaluationQ21 · Clinicians, institutions and societiesQ22 · Re-benchmarking after a modificationQ26 · Agentic devices
Across the five cross-cutting questions
RecovryAI’s reading of the whole filing. Silence is never counted as opposition.
Patient-facing autonomyShould FDA permit patient-facing AI to act with meaningful autonomy within a defined scope?
Supports with conditions
Proportionate evidenceShould evidence requirements scale with clinical risk rather than a uniform high bar?
No position stated
Postmarket relianceCan strong postmarket monitoring justify accepting more premarket uncertainty?
No position stated
Competency evaluationCan a device be evaluated on competency benchmarks and clinical confirmation against clinicians?
No position stated
Change controlCan devices on third-party foundation models be maintained under pre-specified change control?
Supports with conditions
Autonomy acceptedThe highest level this filing accepts
Low-consequence work: Informs
High-consequence work: Directs
This filing was submitted by a member of the public. We publish our analysis here and link to the submission rather than reproducing it, and this page is not indexed by search engines.
Read the full submission on regulations.gov ↗