H.R. 9439 House Bill 119th Congress

Voluntary Consumer AI Disclosure Pilot Act (H.R.9439) — amending the NIST Act to test voluntary approaches first, informed by consumer product labeling that already exists

U.S. House Latest update Jun 24, 2026

A House bill amending the NIST Act to establish a pilot program evaluating voluntary approaches to provenance and disclosure of content produced or manipulated by AI. The text says the pilot is to be informed by consumer product labeling programs already in existence, and that its results support guidelines, best practices and voluntary standards.

Bill overview (primary data)

  • Bill numberH.R. 9439
  • TypeHouse Bill
  • Congress119th Congress
  • Latest actionReferred to the House Committee on Science, Space, and Technology.(2026-06-24)

Key points

  • Amends section 22A of the NIST Act to establish a pilot evaluating voluntary approaches to provenance and disclosure of AI-generated content.
  • Rather than settling a requirement first, it tests voluntary approaches and supports guidelines, best practices and voluntary standards from the results.
  • The pilot is to be informed by consumer product labeling programs already in existence.
  • Two things are covered: content generated by an AI system, and the fact of direct, real-time interaction with one.
  • A report on feasibility, usefulness and limitations goes to Congress within 180 days of completing the pilot.
  • A voluntary approach is piloted before any mandate, with guidance and standards following and existing product labelling taken into account.

1Testing before requiring

Several bills approach disclosure of AI-generated content by requiring it. This one takes a different order: try voluntary approaches first, and build guidelines and standards from what the trial shows.

NIST runs the pilot, and what it evaluates is how private entities might disclose of their own accord. Rather than fixing in advance what must be shown, a stage is inserted to find out which disclosures actually work.

2Informed by consumer product labeling that exists

The text says the pilot is to be informed by consumer product labeling programs in existence at the time it is established. Arrangements for conveying information to consumers, from nutrition labels to energy ratings, are already part of the landscape.

Rather than designing AI disclosure from scratch as a novel problem, the drafting brings in experience from labeling that already runs. What gets read and what gets skipped past is not a question peculiar to AI.

3Two distinct things to disclose

The pilot covers two things. One is content generated by an AI system that a user may engage with. The other is the fact that a direct, real-time interaction with such a system is taking place.

The first is a question of the provenance of an artefact; the second is disclosure that the counterpart is a machine. They tend to be spoken of together as AI disclosure, yet what has to be conveyed, and when, differs. Within 180 days of completing the pilot a report on feasibility, usefulness and limitations goes to Congress.

Of the 114 AI-related bills this site holds as of 2026-09-01, among the 50 not yet covered, 12 amend an existing statute, and this is the one amending the NIST Act.

4Trying before mandating

Several bills take the form of mandating disclosure for AI-generated content. This one takes a different order.

Mandating firstTrying a voluntary approach first (this bill)
Decides in advance what must be shownConfirms which disclosure actually works
Duties attach the moment it passesGuidance and standards follow the results of the pilot
Designs the disclosure from scratchTakes account of consumer product labelling regimes already in existence

Mechanisms for conveying information to consumers already exist in society, from nutrition labels to energy efficiency ratings. What gets read and what gets skipped is not a question peculiar to AI. Two things are evaluated: AI-generated content a user may encounter, and the fact that a direct real-time interaction with AI is taking place.

The first concerns provenance of an output, the second disclosure that the counterpart is a machine.

Why it matters

Designing disclosure is not settled by making it mandatory. What gets read and what misleads is partly unknowable without trying. Inserting a pilot costs time but establishes a form that works before it is fixed. For a business, the voluntary stage is where its own disclosure practice may come under evaluation.

FAQ

Would disclosure become mandatory?
What the bill sets up is a pilot evaluating voluntary approaches. It supports guidelines, best practices and voluntary standards rather than imposing a requirement.
Why draw on consumer product labeling?
The text gives no reason but provides that the pilot be informed by consumer product labeling programs in existence at the time it is established.
What is being disclosed?
Content generated by an AI system, and the fact that a user is interacting directly and in real time with one.

Sources (primary)

Source: Congress.gov (Library of Congress; U.S. legislative materials, public domain). Links go to the official site.

#Congress#Bills#AI#Disclosure#NIST
Disclaimer: This site independently summarizes and classifies information based on official data sources. Always verify the latest and accurate information with the official sources. Content on finance, health, legal, and security is information, not advice. This site is not an official website of the U.S. government.