Before Jonathan Haidt published The Anxious Generation, employees at Meta were discussing something more unusual than a rebuttal to his research: how to persuade other academics to turn against him.
Among the phrases recorded in Meta’s March 2023 internal meeting notes were “Whisper strategy” and “how do we get the academic community to disavow Haidt.” The notes came to light on October 23, 2026, in federal litigation over alleged harm to young people from social media. That is a remarkable glimpse into a company’s response to a critic. It is not, by itself, proof that the company launched a covert campaign or that any particular scholar acted on Meta’s behalf.
The difference matters. So does the underlying paperwork: Exhibit E, containing the internal “Well-being Meeting Sound-check Notes”, and the plaintiffs’ October 23 court reply, which explains why the records were filed.
What the March 2023 Notes Actually Describe
The internal notes span several meetings. On March 7, 2023, staff discussed a response to Haidt’s writing about social media and adolescents. They considered mobilizing researchers who could challenge his methods and getting those objections into news coverage. The notes also anticipated a legal showdown: multidistrict litigation was expected in the coming years, and Haidt’s arguments could feature in it.
By March 8, the discussion had sharpened. Notes attributed a goal to “Curtiss,” apparently Meta’s then-vice-president of research, Curtiss Cobb: getting academics to disavow Haidt. Another entry described a quiet, below-the-radar approach rather than a loud public campaign. The goal appears to have been not merely to respond to the argument but to affect how reporters and other scholars assessed the person making it.
That distinction is at the heart of the story. A company challenging a research finding with evidence is ordinary scientific and public debate. Trying to manufacture peer rejection in order to damage a researcher’s credibility is a different proposition. The internal record suggests employees contemplated the latter. It does not establish who approved such a plan, what budget existed, whether anyone was recruited, or whether the plan was executed.
Contemporaneous reporting by The Verge notes that the pages are rough shorthand, including typos and questions from participants. That makes them important evidence of discussion, but a poor basis for declaring a completed operation.
One Name in the Notes—And a Direct Denial
The notes proposed getting someone called “Gellman” to write a critical piece. That appears to point toward Andrew Gelman, a Columbia University statistician who has openly criticized some of Haidt’s work. But the filing does not independently establish whom the note meant, let alone that the person agreed.
Gelman addressed the story directly on October 27, 2026. In a post on his university-hosted statistics blog, he said he had never been approached by Facebook or Meta regarding Haidt or his research. He also disclosed that Facebook had previously provided funding for unrelated research, including a no-strings-attached donation in 2016. None of that establishes a paid attack on Haidt, and the available record should not be used to insinuate otherwise.
In fact, Gelman’s denial provides an essential test of the bigger allegation. A brainstormed name on an internal document is not evidence that a scholar was enlisted, compromised, or even contacted.
The Book Came Later. The Academic Disagreement Is Real.
Haidt published The Anxious Generation in 2024, after these March 2023 discussions. The book argues that a shift toward phone-based childhood, together with reduced independence and play, has contributed to worsening youth mental health. It became influential in debates over schools, smartphones, platform design, and age restrictions.
That argument has genuine supporters and genuine critics. In a March 2024 review published by Nature, developmental psychologist Candice Odgers disputed whether the evidence supports Haidt’s strongest causal claims. Her critique is on the scientific record; no evidence in the documents discussed here ties her review to Meta’s proposed campaign. It would be just as irresponsible to label all dissent a corporate smear as it would be to ignore the possibility of coordinated influence.
Scientific guidance itself is more nuanced than an all-good or all-bad verdict. The American Psychological Association’s 2023 advisory says outcomes vary with a young person’s vulnerabilities, activities, content and circumstances; it calls for safer platform design and better independent research. The U.S. Surgeon General’s 2023 advisory warns of potential harms while acknowledging the evidence does not establish that social media is sufficiently safe for children.
Whether Haidt is right about every mechanism and every proposed policy is a question for evidence, replication and open criticism—not for corporate image management disguised as academic consensus.
Why These Notes Suddenly Became Public
The papers surfaced through In re Social Media Adolescent Addiction/Personal Injury Products Liability Litigation, federal multidistrict litigation in the Northern District of California, case 4:22-md-03047-YGR. They were introduced during a dispute over documents Meta had withheld or redacted on legal-privilege grounds.
In their October 23 reply (ECF 3604-1), attorneys for plaintiffs including school districts argued that Meta withdrew or narrowed privilege claims on well over half of 300 specially selected documents. They pointed to the well-being notes as an example of material they said was largely public-relations planning rather than protected legal advice. Those are the plaintiffs’ assertions in contested litigation, not a court finding that every challenged document was wrongly concealed. The group of 300 was deliberately selected, not a random sample of all documents.
There is a second thread in the filing: plaintiffs also alleged that some internal youth-well-being research had been delayed, altered or blocked amid legal concerns. Those allegations are important but require their own documentary analysis; they do not prove that the Haidt strategy was carried out.
What the “Whisper strategy” notes provide is narrower and unusually concrete: a view of how employees discussed a critic while worrying about research, reputation and coming litigation.
What Meta and Haidt Have Said—and What Remains Unproved
In comments reported by The Verge, Haidt said he was unsurprised that the company had considered discrediting him. The Guardian reported on October 29 that Meta had been approached for comment; reporting available for this article did not establish an on-record Meta response to the specific notes.
The unanswered questions are straightforward: Did Meta authorize a covert campaign? Did it approach any academics or journalists? Did it place or commission criticism without disclosing its role? The released meeting notes do not answer those questions. Nor do they prove that criticism of Haidt’s scholarship is illegitimate.
But the questions deserve answers. Companies have every right to contest research that affects their products. If a business chooses instead to quietly shape who gets treated as a credible researcher, the public may never know whose judgment it is really hearing.
The question isn’t whether Jonathan Haidt should face scrutiny. It’s whether the scrutiny we see is independent—or whether someone with a financial stake helped write the script.
Watch and Read the Evidence
Video commentary: Watch the short about Meta and The Anxious Generation that prompted this examination. It is commentary on the reported filings, not original court footage or proof that a campaign occurred.
Contextual interview: PBS’s April 1, 2024 interview with Jonathan Haidt explains his argument about smartphones and youth mental health. The interview predates the release of the internal Meta notes.
- Primary record: Exhibit E, Meta’s internal well-being meeting notes (ECF 3604-6)
- Primary record: plaintiffs’ reply filed October 23, 2026 (ECF 3604-1)
- Andrew Gelman’s direct response, October 27, 2026
- Candice Odgers’s published critique of The Anxious Generation
Featured photograph: actual Meta headquarters in Menlo Park, photographed by Nokia621 (Wikimedia Commons, CC BY-SA 4.0). The photo is contextual; it is not the court exhibits or internal meeting notes. For related Apep Bar reporting on Meta and privacy, see the company’s smart-glasses recognition plans and Meta AI link-sharing exposure.
