OpenAI CEO Briefs House Judiciary on Automated Copyright Issues
OpenAI CEO Sam Altman briefed House Judiciary Committee members Tuesday, Chairman Jerry Nadler, D-N.Y., said as members grapple with potential solutions for crediting copyright owners when artificial intelligence systems use their work.
Sign up for a free preview to unlock the rest of this article
Communications Daily is required reading for senior executives at top telecom corporations, law firms, lobbying organizations, associations and government agencies (including the FCC). Join them today!
Altman testified before Senate Judiciary Committee members Tuesday (see 2305160074), opining on copyright and election disinformation associated with AI. The House Intellectual Property Subcommittee focused Wednesday’s hearing on finding a balance between promoting AI innovation and protecting the works of musicians, photographers, writers and other creative professionals.
It’s not an easy balance to strike, said Nadler. He cited Copyright Office guidance saying only works of “human authorship” are eligible for copyright protection. However, Congress needs to draw the line where human input results in automated content that should be protected.
Spain is moving forward with restrictive regulations for AI copyright, while Japanese lawmakers believe all information that goes into AI teaching should be free of copyright restrictions, said House Intellectual Property Subcommittee Chairman Darrell Issa, R-Calif. Congress will need to find a middle ground between the two approaches, he said.
“There are significant harms caused to the creative community when AI developers ignore copyright law and allow their AI programs to ingest copyrighted works without obtaining permission from -- or compensating -- the copyright owners of the works they use,” said Copyright Alliance CEO Keith Kupferschmid in a statement on Wednesday’s hearing. He noted there are existing licensing systems for literary works and visual art that could be useful in the AI conversation, and credited companies for recently recognizing copyright issues for automation. Both Altman and IBM Chief Privacy and Trust Officer Christina Montgomery noted the issue during Tuesday’s hearing. Nadler said during Wednesday’s hearing OpenAI is valued at about $29 billion, and asked witnesses for salary estimates in various creative professions. PLUS Coalition CEO Jeff Sedlik, a professional photographer, told Nadler photographers can usually make $20,000-$40,000 per year, though it’s possible to make more. Society of Composers and Lyricists President Ashley Irwin and singer-songwriter Dan Navarro said earnings are wide-ranging in their professions.
Congress is going to need to deal with this gray area that melds human creation with automation, said Latham & Watkins attorney Sy Damle, former general counsel for the Copyright Office. When a human is exercising control over a creation, there should be copyrightable output to some degree, he said: AI use is comparable to a photographer using the latest camera and digital technology.
AI can scrape copyright material from the internet and compose something original fairly quickly, while a musician might take several years to write certain songs, said ranking member Hank Johnson, D-Ga. Humans are the only creators with the “discerning taste” to make creative works successful, said Irwin. Issa suggested there should be a way to enforce copyright protection through database inputs searchable by users. Rep. Zoe Lofgren, D-Calif., asked if it’s possible to reach some sort of agreement between various parties the way Congress did with the Music Modernization Act, which dealt with royalties for online streaming (see 2208150042). The practical implementation for works created by AI won’t be so straightforward because AI can compile a number of different works and distill them into something that’s essentially original, said University of Pennsylvania associate professor Chris Callison-Burch.