“Singling out one specific technology for disparate treatment would set a dangerous precedent for other new inventions and technologies in the future. – C4IP statement
As the U.S. Senate disbanded from its 24-hour “vote-a-rama” debating amendments to President Donald Trump’s “One Big Beautiful Bill Act” (OBBBA) earlier today, one major provision that would have impacted the regulation of artificial intelligence has been scrapped.
The AI modernization provision of the original version of the bill would have banned state and local governments from regulating AI for 10 years if they wanted access to funds aimed at improving AI infrastructure. Senator Marsha Blackburn (R-TN) initially joined a compromise amendment with Senator Ted Cruz (R-TX) that would have narrowed the scope of the provision and lowered the moratorium to five years, but later withdrew her support for that amendment and introduced her proposal to strike the provision in its entirety along with Senator Maria Cantwell (D-WA), which passed by a vote of 99-1. Senator Thom Tillis (R-NC), who recently announced he will not be running for another term, was the sole no vote. Tillis also joined two other Republican senators in voting against the bill overall.
Separately, Tillis proposed his own amendment to the bill, titled the “Tackling Predatory Litigation Funding Act,” which did not pass. US Inventor called it a “victory” and said the amendment “threatened to severely restrict litigation funding—one of the few avenues available for American inventors and startups to defend themselves against large corporations attempting to steal their technologies. Its passage would have strengthened Big Tech monopolies and enabled China to advance further in critical technologies.”
Tillis’ decision to retire could also have some effect on passage of the patent bills currently pending in Congress—the Patent Eligibility Restoration Act of 2025 and the Promoting and Respecting Economically Vital American Innovation Leadership (PREVAIL) Act
Prior to the AI amendment, Frank Cullen, executive director of the Council for Innovation Promotion (C4IP), released a statement urging the senators to be cautious about a full moratorium on regulation. The statement said that while C4IP “recognizes and appreciates lawmakers’ good intentions behind the proposed moratorium… a moratorium threatens to do more harm than good.” Cullen added:
“States are the ‘laboratories of democracy,’ as Supreme Court Justice Louis Brandeis famously noted. And state-level regulatory frameworks have historically added important protections to intellectual property rights and consumer interests. Foreclosing these opportunities to protect creators and consumers would be a significant loss to the United States. In addition, singling out one specific technology for disparate treatment would set a dangerous precedent for other new inventions and technologies in the future.
On balance, this moratorium will be harmful to intellectual property. As a result, C4IP will score it negatively in its annual Congressional Innovation Scorecard.”
And today, the Human Artistry Campaign said it was grateful to Blackburn and Cantwell and applauded the amendment. “This dangerous and unwise proposal would have banned states from passing new legislation or enforcing existing laws, including to provide greater transparency about whose art was copied to create AI models and banning nonconsensual deepfakes and voice clones. There is a better way forward with policies that recognize AI could be a valuable tool aiding human creativity, but should never be used to destroy artistry and culture.”
The Recording Industry Association of America (RIAA) also cheered the decision to scrap the moratorium, noting that “[t]his final vote is a resounding dismissal of the shortsighted, zero-sum game thinking from some tech companies and signifies a widespread recognition that respect for human creativity does not impede winning the AI race, but instead is essential to its success.”
The bill will now return to the House of Representatives for reconciliation, but it still faces opposition from some key Republicans.
For more on the IP aspects of the OBBBA, read here.
Image Source: Deposit Photos
Image ID:84022156
Copyright:gustavofrazao
Join the Discussion
4 comments so far.
Anon
July 2, 2025 10:59 amAs for:
“As for asking AI to interpret the [nearly indecipherable] MPEP?
‘I’ve asked it questions about the MPEP and it says that certain concepts are supported in certain sections, but it’s just not there. I have to circle back and give it feedback and tell it not to do that again but it still does it but you have to be aware enough to know when it is BSing.’
Source reddit/patentexaminer thread”
Most Gen AI engines are built to be obsequious, and there is not a single one that I have dabbled with that I cannot redirect its output to give me an opposite conclusion from an initial provision (and I have switched both from incorrect to correct and from correct to incorrect).
And this is besides the rather uncontrolled possible use by examiners of unapproved Gen AI engines, given the extensive nature of remote work.
Julie Burke
July 2, 2025 08:16 amAs for AI-usage inside the USPTO, patent examiners raise concerns about high AI-error rates when used for menial office tasks.
Even ChatGTP is aware of its own shortcomings.
“ChatGPT: “AI systems—especially in open-ended tasks like search, summarization, scientific explanation, multi-step automation, or news reporting—frequently exhibit error rates well over 50%, sometimes even approaching 80–90%. These assessments are rooted in peer-reviewed research and credible audits, not just anecdotes.””
As for asking AI to interpret the [nearly indecipherable] MPEP?
“I’ve asked it questions about the MPEP and it says that certain concepts are supported in certain sections, but it’s just not there. I have to circle back and give it feedback and tell it not to do that again but it still does it but you have to be aware enough to know when it is BSing.”
Source reddit/patentexaminer thread
francis hagel
July 2, 2025 02:22 amOn the cognitive risks of intensive use of LLM for essay writing, see MIT Media’s study available at
https://www.media.mit.edu/publications/your-brain-on-chatgpt/
This should be a warning for the legal and IP professions. Think twice before expanding the use of ChatGPT and the like to reduce headcount.
Anon
July 1, 2025 03:46 pm99-1….
That tells me that the issue is simply not important to Congress.
Do we really want any controlling legisation on AI to be of the haphazard State by State laws that already govern rights of publicity and privacy?