A new bill in the New York state legislature would require news organizations to label AI-generated material and mandate that humans review any such content before publication. On Monday, Senator Patricia Fahy (D-Albany) and Assemblymember Nily Rozic (D-NYC) introduced the bill, called The New York Fundamental Artificial Intelligence Requirements in News Act — The NY FAIR News Act for short.

“At the center of the news industry, New York has a strong interest in preserving journalism and protecting the workers who produce it,” said Rozic in a statement announcing the bill.

A closer look at the bill shows a few regulations, mostly centered around AI transparency, both for the public and in the newsroom. For one, the law would demand that news organizations put disclaimers on any published content that is “substantially composed, authored, or created through the use of generative artificial intelligence.”

AI disclaimers for readers have been hotly debated in the news industry, with some critics arguing that such labels alienate audiences, even when generative AI is only used as an assistive tool. The bill contains a carve-out that would allow copyrightable material to be excluded from the law. (The U.S. Copyright Office has ruled that works solely generated by AI systems are not eligible for copyright, but allows leeway for works that show signs of “human authorship.”)

The bill also requires that news organizations disclose to journalists and other media professionals in their newsrooms when AI is being used and how. Any news content created using generative AI must also be reviewed by a human employee “with editorial control” before publication. That goes not just for news articles but also for audio, images, and other visuals.

In addition, the bill contains language that requires news organizations to create safeguards that protect confidential material — mainly, information about sources — from being accessed by AI technologies.

State lawmakers highlighted two main reasons for proposing the NY FAIR News Act. First, they say, AI-generated content may be “false or misleading.” Second, they argue, AI-generated content “plagiarizes” by deriving content from original sources “without permission or proper citation.”

“Perhaps one of the industries at most risk from the use of artificial intelligence is journalism and as a result, the public’s trust and confidence in accurate news reporting,” said Sen. Fahy in a statement. “More than 76% of Americans are concerned about AI stealing or reproducing journalism and local news stories.”

The proposed bill was announced with broad endorsements from unions across the news industry, including WGA-East, SAG-AFTRA and the DGA.

Jennifer Sheehan, a spokesperson for the NewsGuild of New York, confirmed that the NewsGuild has been meeting with this labor coalition to discuss shared concerns around AI adoption and working to get the bill off the ground.

Notably, the bill would cement some labor protections for newsroom workers — including restrictions on firing journalists or reducing their work, pay, or benefits due to generative AI adoption. Similar language has been negotiated into individual newsroom union contracts across the country over the past couple of years.

In December, the NewsGuild launched a nationwide campaign called “News Not Slop” to advocate for more guardrails on AI usage in newsrooms. In New York City, the Business Insider union held a rally in the Financial District to protest an editorial pilot that was publishing AI-generated news stories with an “AI byline.”

“Our union is deeply concerned about media companies implementing artificial intelligence in ways that damage the credibility of our members’ journalism,” Sheehan said, “as well as the impact such technology has had and will have on jobs.”