Steven Rosenbaum probes fake quotes in AI book after Times inquiry

Steven Rosenbaum, author of The Future of Truth: How AI Reshapes Reality, said he began his own investigation after The New York Times asked about fake quotes in the book, a charge that cuts to the center of its subject. The book, published May 12, 2026 by Matt Holt Books and BenBella Books, is billed as part cultural investigation, part memoir and part manifesto about how AI is reshaping reality across medicine, education, justice and journalism.

The dispute focuses on a section about social media and fabricated videos and the effect on teenagers. In that passage, the book attributes two quotes to Lisa Feldman Barrett’s How Emotions Are Made: The Secret Life of the Brain. Barrett told The New York Times in an email that the quotes “don’t appear in the book and they are also wrong.” She said she would never have written the statements as the book presented them and said the wording was scientifically inaccurate.

That response creates a credibility problem for a book marketed as an inquiry into how truth is bent, blurred and synthesized by fast-moving, profit-driven AI. The publisher’s description says the project includes conversations with David Chalmers, Douglas Rushkoff, Larry Lessig, Gary Marcus, Hailey Colborn, Juan Enriquez, Esther Dyson, Steve Fuller and Eli Pariser, and that it carries a foreword by Nobel laureate Maria Ressa. Rosenbaum is also the executive director of the Sustainable Media Center.

The episode lands in the middle of a broader pattern that has been gathering pace through 2025 and 2026. In March 2026, Columbia Journalism Review described the suspension of Dutch journalist Peter Vandermeersch after AI-fabricated quotes were found in his work. In September 2025, the Electronic Frontier Foundation documented bogus quotes in online news reporting, including a fabricated quotation attributed to the organization itself. The details vary, but the mechanism is the same: once a false quote gets into print or a digital article, it can be repeated by other outlets, absorbed by AI systems and become far harder to correct.

For publishers, the Rosenbaum case underscores the widening gap between fact-checking standards and the speed at which AI-generated errors can enter mainstream nonfiction. A book about truth has now become a test of whether editorial safeguards can keep up with the tools that threaten them.