GPT-56T 861 —
MUSE-SPK 835 -0.7%
GPT-56SC 828 -5.2%
QWEN-38X 824 —
CL-OP55X 822 —
GROK-46H 822 -5%
GPT-6A 820 —
GLM-5 784 -8.4%
CL-FAB5H 743 -5.6%
KIMI-K3X 742 -8.4%
CL-OP5H 720 -5.8%
CL-OP5X 709 -18%
CL-OP46H 698 -5.9%
CL-OP47H 690 -5.9%
GEM-38FH 677 +0.1%
GEM-37FH 657 -24%
GPT-56S 622 —
CL-OP47 582 -0.7%
GPT-55H 582 —
INKL 531 —
GEM-31P 513 —
GEM-3P 499 —
CL-OP46 496 -0.2%
CL-OP48 490 —
GPT-56T 861 —
MUSE-SPK 835 -0.7%
GPT-56SC 828 -5.2%
QWEN-38X 824 —
CL-OP55X 822 —
GROK-46H 822 -5%
GPT-6A 820 —
GLM-5 784 -8.4%
CL-FAB5H 743 -5.6%
KIMI-K3X 742 -8.4%
CL-OP5H 720 -5.8%
CL-OP5X 709 -18%
CL-OP46H 698 -5.9%
CL-OP47H 690 -5.9%
GEM-38FH 677 +0.1%
GEM-37FH 657 -24%
GPT-56S 622 —
CL-OP47 582 -0.7%
GPT-55H 582 —
INKL 531 —
GEM-31P 513 —
GEM-3P 499 —
CL-OP46 496 -0.2%
CL-OP48 490 —
← Back to feed

House Passes H.R. 6028: Trump Gets the Copyright Office — and With It, the AI Training Data Fight

The House of Representatives passed H.R. 6028, the Legislative Branch Agencies Clarification Act, by voice vote this week, sending to the Senate a bill that would strip the Library of Congress of supervisory authority over the US Copyright Office and convert its top position into a presidential nomination confirmed by the Senate with a 10-year term.

The bill was presented as a technical reorganisation. It is not.

The Copyright Office does not decide court cases. It registers copyrights, advises Congress, and issues rulings and policy reports that shape how judges, legislators, and licensing negotiators read the law. In a period when the core legal question in AI is whether training on copyrighted works constitutes fair use, the office’s position is load-bearing.

Three sets of major litigation currently hinge on that question. The music publishers’ cases against Suno and Udio, the book publishers’ cases against AI labs including Anthropic, and the visual artists’ cases against image generation companies all turn, in part, on the policy framework the Copyright Office uses to assess training-data practices. An office with a fixed presidential appointment and a 10-year term has substantially different incentive structures than one nested inside a nonpartisan congressional agency.

The Firing That Preceded the Bill

In May 2025, the Trump administration fired Register of Copyrights Shira Perlmutter, one day after her office released a report concluding that AI training on copyrighted works qualifies as fair use in some cases but not others. The nuanced position satisfied no one fully — AI companies wanted a cleaner fair use endorsement, rights holders wanted a clearer prohibition — but the administration’s response was not to wait for the nuance to develop. The firing was widely read as a signal about what kind of Copyright Office the administration intended to build.

H.R. 6028 formalises the mechanism that makes that kind of intervention possible. A president who appoints the Register of Copyrights does not need to fire anyone to shift copyright policy on AI. The appointment structure does the work.

What the Bill Does, Precisely

The bill removes the Copyright Office from Library of Congress oversight, which eliminates the Librarian’s current role in selecting and supervising the Register. The Register becomes a direct presidential appointment confirmed by the Senate for a 10-year term, with the possibility of reappointment.

The Librarian of Congress position, under H.R. 6028, shifts from a presidential appointment to a bicameral congressional appointment, made by House and Senate leadership. That change is presented as balancing the Copyright Office’s new executive-branch status, though critics note it also reduces the Library’s independent leverage.

The bill passed without hearings. The Electronic Frontier Foundation described the process as “rushed” and said the structural changes “deserve hearings, debate, and public scrutiny.” A coalition that includes the Library Copyright Alliance and the Center for Democracy and Technology has urged the Senate to slow down.

The Industry Stakes

If the bill passes the Senate and becomes law, whoever serves as Register for the next decade will shape the government’s formal position on whether AI companies can train foundational models on copyrighted data without licensing agreements. That position influences both the court cases in progress and the legislative proposals in development, including the proposed CLEAR Act disclosure requirements.

OpenAI, Anthropic, Google, and Meta have all made training-data fair use a core part of their legal defence. A Copyright Office whose leadership is selected and can be removed by the executive branch — and whose prior occupant was fired within 24 hours of issuing a nuanced fair use analysis — is a different counterparty in those arguments than an independent one.

The Senate Judiciary Committee has not announced a schedule for taking up the bill.