Former Official Urges IP Protection in AI Development

Coke Morgan Stewart calls on the U.S. to defend creator rights amid ongoing legal battles over AI training.

Updated on Oct. 27, 2026 in Artificial Intelligence

Bold flat-color editorial illustration of a heavy monolithic seal, symbolizing the structure of intellectual property protection in the digital era.
Former USPTO official Coke Morgan Stewart is urging the federal government to strengthen intellectual property protections amidst ongoing copyright litigation involving OpenAI. AI Illustration. Upload story photo >

Live Poll

Should AI companies be required to obtain permission before training models on copyrighted works?

Former USPTO deputy director Coke Morgan Stewart is pushing the federal government to prioritize intellectual property protections as AI development accelerates. Her appeal comes as the U.S. navigates high-stakes copyright litigation involving companies like OpenAI.

Why it matters

Stewart argues that robust intellectual property rights act as a critical economic equalizer for creators in an era of rapid technological advancement. Ensuring these rights remain protected is a central tension in current disputes over whether training large language models on copyrighted content is permissible.

Startups paired with more lenient patent examiners saw 55% higher employment and 80% higher sales growth five years later. Research into this correlation highlights how regulatory oversight impacts business performance across the innovation sector.

The players

Coke Morgan Stewart

She is the president and CEO of the Council for Innovation Promotion and served as the deputy director of the U.S. Patent and Trademark Office until July 2026.

OpenAI

This AI research and deployment company is the defendant in consolidated litigation regarding the use of copyrighted text for model training.

The New York Times

This prominent media organization is among the publishers currently alleging copyright infringement against AI developers.

United States Patent and Trademark Office

This federal agency is responsible for granting patents and registering trademarks to promote industrial and technological progress.

The details

The United States recently submitted a 20-page Statement of Interest in a consolidated case against OpenAI, maintaining that training AI on copyrighted material is not inherently unlawful. The government's position is being weighed against claims from plaintiffs like The New York Times, which allege widespread copyright infringement by AI developers.

Timeline

  1. Coke Morgan Stewart concluded her term as deputy director of the USPTO in July 2026.

  2. The United States filed its Statement of Interest in the OpenAI litigation on Sept. 1, 2026.

  3. Stewart publicly urged the government to preserve creator protections on Oct. 27, 2026.

  4. Replies regarding summary-judgment motions in the consolidated case are due Nov. 20, 2026.

The Tech Race

This dispute marks a critical test for the Federal Copyright Act as it faces pressure to adapt to modern generative AI technologies. The outcome will shape the arms race between tech developers and content creators over the ownership of training data.

Users may see changes in the availability of AI-generated content or subscription costs if copyright settlements force developers to pay for training data. These shifts could ultimately dictate which tools and media platforms remain viable for daily use.

The takeaway

The intersection of artificial intelligence and intellectual property law remains in flux as courts define the scope of fair use. Readers should watch for upcoming summary-judgment rulings, as these will likely set the rules for how AI tools are built and what content they can access.

What happens next

The court will receive replies on summary-judgment motions in the consolidated OpenAI litigation on Nov. 20, 2026.

Further reading

For more information on the evolving standards for machine learning, visit the Artificial Intelligence section.

Live Poll

Should AI companies be required to obtain permission before training models on copyrighted works?