A campaign opposing Waymo’s robotaxis used AI in opinion pieces published under human authors’ names, Semafor reports. The case puts a practical question on the table for readers and publishers: when advocacy copy gets AI assistance, who should disclose it, and how much?
Watch Desk analysis
What happened
Semafor says AI-detection service Pangram found significant AI input in several opinion pieces criticising autonomous vehicles, including articles published by the Dallas Morning News, the San Francisco Chronicle and the New York Daily News. Pangram’s percentages are its estimates, not independent proof of who wrote each passage or how much AI contributed.
The report says some pieces were drafted by a public-relations firm. Ira Goldstein, whose New York Daily News op-ed was among them, told Semafor the firm prepared a first draft and he reviewed and edited it. He said he stood by the piece. A spokesperson for lawyer Amy Witherite likewise said her firm develops and drafts pieces, sometimes using AI, and that staff review and fact-check them before approval.
The articles argued against autonomous vehicles, citing concerns including jobs, safety, access to government information and blocked emergency vehicles. Semafor also reports that the campaign included paid advertising and involved opponents of Waymo. That is a broader campaign than a few stray paragraphs getting an algorithmic polish.
Why it matters
Opinion pages are meant to carry an identifiable person’s argument. Ghostwriting and PR assistance are not new, but AI tools make the line between help with a draft and substantial machine-generated text harder to see. Readers may reasonably want to know how an argument was made, particularly when it is part of organised advocacy.
The report also captures a real disagreement about disclosure. Some writers and publishers see AI assistance as routine; others say readers deserve original opinions and clearer guardrails. Detection scores alone cannot settle that argument, and they should not be treated as a verdict on an individual author.
Our read
The useful point is not that an AI detector has caught a campaign red-handed. It is that advocacy can travel through familiar local opinion pages while the role of PR firms and AI remains unclear to readers. Disclosure rules should distinguish light editing from a machine-assisted first draft, rather than pretending every use of a writing tool is the same.
What to watch
- Whether the publications clarify what authors must disclose about AI and outside drafting help.
- Whether campaign groups and PR firms adopt clearer disclosure practices.
- How editors handle detector results without treating estimates as proof.
Discussion spark: Should opinion pages require disclosure whenever AI helps draft a piece, or only when it writes a substantial part of the argument?
Sources and evidence
- Anti-Waymo campaign relied on op-eds written with AI | Semafor (9 November 2026, 02:06 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.