355
Audio & Video Production344
Automation & Workflow224
Software Development250
Marketing & Growth192
AI Infrastructure & MLOps173
Writing & Content Creation203
Data & Analytics140
Design & Creative169
Customer Support130
Photography & Imaging156
Sales & Outreach125
Voice & Speech135
Operations & Admin87
Education & Learning131
WIRED pulled an excerpt from Steve Rosenbaum’s book after reports of fake quotes and concerns about how much AI was used in the writing process.
In short: WIRED removed a published excerpt from Steve Rosenbaum’s book after concerns about whether parts were written or edited with AI.
WIRED published an excerpt from Steve Rosenbaum’s book, The Future of Truth, which looks at how artificial intelligence can affect what people believe is real. Soon after, The New York Times reported the book included more than a half-dozen quotes that were made up or credited to the wrong people. Rosenbaum later said he had accidentally included “a handful” of quotes that were “improperly attributed or synthetic,” meaning they were not correctly sourced and some were generated.
WIRED then rechecked its 1,450-word excerpt and said its fact-checkers confirmed the quotes and facts in that excerpt were accurate. But WIRED also has a rule against publishing AI-generated or AI-edited writing. After a reader claimed the excerpt looked “blatantly AI-written,” WIRED used several AI detection tools, including Pangram, GPTZero, and ZeroGPT. These tools suggested the excerpt was likely AI-generated, but WIRED noted that such tools can be wrong.
Rosenbaum told WIRED he used AI tools for research, brainstorming, structure, and “language refinement,” but said the excerpt was not generated by AI and published as-is. When WIRED tested the full book with Pangram, it flagged the text as 53 percent AI-generated, plus 9 percent likely AI-assisted. Rosenbaum declined to engage with the percentage results and gave unclear answers on whether he copied and edited AI text.
WIRED said it is revising its AI guidelines, but one rule will stay: published work cannot be written with AI. Due to uncertainty about Rosenbaum’s process, WIRED retracted the excerpt.
This dispute shows a growing problem for readers, publishers, and newsrooms: it can be hard to tell what was written by a person versus “helped” by AI, and even the tools that try to detect AI can misfire (like a smoke alarm that sometimes goes off when nothing is burning).
Source: Wired