A decision matrix is more useful than a rights poster
The matrix tests each proposal against four questions: what law currently requires or permits; what a proposed federal provision would add or replace; which official, sponsor, provider or recipient must act; and what happens if that actor fails. It should also identify implementation cost and the strongest objection. A proposal with no responsible actor or remedy is a declaration, not an operational right.
A useful matrix must distinguish an already existing regulatory feature from an intended reform. Individual sponsor status, some expedited review categories, clinical holds, written sponsor delegation and informed consent duties already exist. Simply announcing those features again without adding access, cost reduction, timing or review would not move the ball forward. F02F06F09
The table on this site presents proposed mechanisms alongside source-level law. It deliberately leaves contested issues as alternatives for legislative counsel. That is not indecision: it is how a responsible congressional office learns what must be resolved before introducing enforceable text.
How to convert this matrix into bill instructions
For each row, legislative counsel should identify an existing section of the U.S. Code, whether amending it is the preferred technique, any implementing regulation, appropriations or enforcement authority required, and its interaction with other laws. Some rows call for new procedural duties, others for funded services or rights against research-data custodians. They should not all be collapsed into a blanket IND exemption. F01F12F16
Agency actions should include the relevant filing definition, clock starts and stops, required reasons for refusal, appeal jurisdiction, and public aggregate reporting. Where private rights are proposed, define standing, available remedies, record-retention duties, any notice-and-cure requirement, and safe harbor for good-faith adherence without erasing negligence or fraud remedies.
A companion research evaluation should track small-sponsor applications, the proportion of complete filings, classification response time, review access and cost, voluntary participant comprehension, adverse-event handling, reported studies, negative results and data withdrawal compliance. Public performance evidence can justify future extension or revision.
Avoid designing new federal law by accident
The federal government cannot automatically turn every private IRB, state medical board or insurer into an agency subject to a new administrative appeals process. A statutory remedy should specify exactly which actor is regulated and under what jurisdiction. Existing agency petition and review procedures should be compared before a new one is invented. F09F10
A research participant's acknowledgment of uncertain effects should not be used to disclaim responsibility for product contamination, negligent conduct, fraudulent claims, or the absence of consent. If Congress wants an injury fund, it must decide whether funding is via appropriation, fee, insurance, or pooled risk mechanism and specify eligible injuries and an appeal path. F07F08
Genomic and de-identified data rights require equally deliberate drafting. HIPAA alone does not guarantee every future downstream AI system will honor a participant's use restriction. Congress could establish a separate purpose-limitation rule, but that is a new normative choice with defined exceptions, feasibility constraints, and enforcement costs. F14F15F17