When journalist Ryan Sabalow from CalMatters began reporting on California’s legislature in 2023, he noticed a curious trend. Lawmakers would stand before the chamber and passionately denounce a bill — only to avoid voting on it altogether. The pattern sparked a question: How many controversial decisions in Sacramento were quietly shaped not by votes, but by silence?

It was a question that once would have required endless hours digging through public records and legislative transcripts. But in an era reshaped by generative artificial intelligence, Sabalow didn’t need to sift through dusty filing cabinets or spreadsheets. He simply turned to Digital Democracy, an advanced A.I. system capable of archiving every spoken word, campaign contribution, and recorded vote in California’s legislative proceedings. The results led to a revealing article and an Emmy-winning CBS segment that exposed how Democratic legislators quietly killed a widely supported fentanyl bill by choosing not to vote at all.

“I don’t think I could have done that without this database,” Sabalow explained.

His experience represents a broader shift happening across the news industry. Artificial intelligence is rapidly transforming how journalists gather information, analyze complex records, and produce content. Legacy media outlets are increasingly adopting tools built by companies such as Google and OpenAI to streamline research, accelerate editorial tasks, and even generate first-draft copy. Functions that once consumed hours — searching archives, cross-checking data, identifying sources, testing story ideas — can now be completed in minutes with machine assistance.

Some publishers, including Business Insider and Fortune, have experimented with A.I.-generated articles, informing audiences that machine tools would contribute to early drafts. Most organizations, however, maintain strict editorial controls. Human editors must review any A.I.-assisted output before publication — a safeguard that has not always prevented mistakes. Even top publications like Bloomberg and Wired have issued corrections for A.I.-related errors.

The enthusiasm for artificial intelligence in media comes paired with anxiety. Newsrooms continue shrinking as advertising revenue declines and digital platforms dominate attention. Many journalists wonder whether A.I. will streamline their work — or replace their jobs entirely.

“A.I. is a remarkable resource for reporters,” said Stephen Adler, former editor in chief of Reuters and now head of New York University’s Ethics and Journalism Initiative. “It can analyze large datasets, help organize information, spot grammar issues, and even flag potential gaps in a story. But it also brings major risks if misused or over-relied upon.”

The industry’s stakes are high. After decades of struggle adapting to the internet age, media companies do not want to repeat past mistakes. Many executives argue that tech giants owe compensation for training A.I. models on news content — an issue driving new licensing deals and lawsuits. For instance, The New York Times has sued OpenAI and Microsoft, alleging that their models improperly used copyrighted journalism to generate answers. Both companies deny wrongdoing.

Despite legal disputes, many organizations are pressing forward with A.I. innovation. Newsquest in the United Kingdom has employed dozens of reporters trained to leverage A.I. tools. Axel Springer, parent company of Politico and Business Insider, built an A.I.-enabled travel planner. Time magazine used machine intelligence to enhance its coverage of the 2024 Person of the Year. The New York Times now has a team dedicated to A.I. experiments and newsroom technology.

Axios stands among the most vocal proponents of artificial intelligence in news production. The outlet uses machine models to help curate its local newsletters and automatically collect top headlines for daily briefings, with human oversight. Allison Murphy, Axios’s chief operating officer, stresses that automation is designed to support — not replace — reporters.

“Anything that isn’t about human judgment, let’s find the fastest and most efficient way to do it,” Murphy said. She noted that Axios journalists can also use a tool called the Axiomizer to draft and refine the outlet’s signature writing elements, such as “Why it matters” context lines.

The Associated Press demonstrates another use case: investigative research at scale. When A.P. reporters analyzed archival files related to the assassinations of President John F. Kennedy, Martin Luther King Jr., and Robert F. Kennedy, A.I. tools helped search and summarize tens of thousands of pages — including newly un-redacted sections. The technology saved journalists days of manual review and enabled faster public analysis.

Not every newsroom welcomes such rapid adoption. At The Washington Post, internal debate erupted when an engineer questioned whether a proposed A.I. tool — designed to summarize and assemble information from other publications — risked copyright violations and ethical pitfalls. Leadership pushed back, the engineer eventually resigned over A.I. ethics concerns, and the company ultimately launched a narrower product for automated headline displays.

Bloomberg has also faced scrutiny after A.I.-generated article summaries led to factual corrections, including mischaracterizing economic data in one case and misattributing quotes in another. The company insists that the vast majority of A.I.-assisted work meets editorial standards and emphasizes transparency when tools are used.

Labor unions have taken a prominent role in shaping A.I. policy in media workplaces. The NewsGuild, representing thousands of journalists across the United States, has negotiated dozens of agreements since late 2023 dealing with automation safeguards and job protections. Union leaders argue that formal rules are required to prevent newsroom automation from eroding editorial roles.

Disputes are already emerging. Journalists at Politico and E&E News accused their parent company of using automated live coverage without notifying the union, triggering an arbitration case. The outlet has not commented publicly on the dispute.

Meanwhile, organizations like NPR are testing new workflows that could allow A.I. to help convert radio broadcasts into written articles. Some employees raised concerns that automating “the middle” of the editorial process could jeopardize vital journalistic judgment.

As newsrooms confront both opportunity and uncertainty, one thing is clear: artificial intelligence is no longer a distant concept. It is already embedded in the editorial process — helping reporters uncover hidden stories, accelerating information analysis, and shaping the future of media production.

Yet, even as machines become more capable, many in the industry believe that the most essential element of journalism remains unchanged: human judgment, curiosity, and integrity. Whether A.I. becomes a tool for empowerment or a source of disruption will depend on how news organizations navigate this pivotal moment — and how committed they remain to the values that define trustworthy reporting.