Five years ago Frances Haugen’s documents revealed what Facebook, and Instagram, knew about the effects of its social media on users and what it had chosen to keep quiet. Five years later, it’s worth asking what has really changed.
The data scientist Frances Haugen handed over thousands of internal documents to the Wall Street Journal in 2021 Facebook. From those papers it emerged that the company knew perfectly which ones damage its platforms (including Instagram) were provoking the safety and to mental health of users, but chose not to intervene. Solving those critical issues, in fact, would have entailed a reducing the time people spend on apps. The files stolen by Haugen ended up at the center of thousands of lawsuits on the damage caused to young people, proceedings which forced Meta to pay billion dollars and to review some features of their social networks. Today, five years after the release of those documents, it is legitimate to ask whether that same algorithm at the center of the investigation, which rewarded anger and polarization, still works in the same way.
It is precisely this enormous scandal that is at the heart of The Social – The price of truth. Sixteen years later The Social Networkdirector Aaron Sorkin returns to shine the spotlight on Mark Zuckerberg’s company, bringing the figure of Frances Haugen and the consequences of her revelations to the big screen.
What was in the documents: The Facebook Files
Frances Haugen was part of the team Facebook Civic Integritythe group responsible for identifying and stemming misinformation on the platform, with particular attention to that which circulated during the electoral rounds. When the company disbanded that group, she decided to leave, taking with her pages and pages of internal research to give to reporter Jeff Horwitz of the Wall Street Journal, which would later become a series of investigations known as The Facebook Files.
Haugen later filed a complaint with the SEC, the American securities regulator, and testified before the United States Senate, declaring that “Facebook’s products harm children, fuel division, and weaken our democracy. The company’s leaders know how to make Facebook and Instagram safer, but they are not making the necessary changes because they have put their astronomical profits before people.” Furthermore, in his speech, Haugen specified that the documents provided to Congress “prove that Facebook has repeatedly misled the public about what its own research reveals about the safety of children, the effectiveness of its artificial intelligence systems, and its role in spreading divisive and extreme messages”.
As can be read in the investigation, “32% of teenagers said that when they felt dissatisfied with their body, Instagram made them feel even worse“. This is what emerged in a March 2020 presentation, published on the company’s internal bulletin board and viewed by the WSJ. In the same period, the article says, Instagram’s internal researchers were studying experiences of this type, wondering if they were part of a broader phenomenon, understanding in short how Instagram was “harmful for a significant percentage of these users, especially for teenage girls”.
Furthermore, internal research commissioned by the company had documented that the automatic systems intercepted, according to the estimates cited by Haugen, only 3-5% of hateful content and just 0.6% in the case of violent content or content that incited violence. For Haugen the crux lies in the algorithm introduced by Facebook in 2018built to push engagement and to prioritize Meaningful Social Interactions, i.e. significant social interactions, with the aim of encouraging real exchanges between friends and family on the platform. The result was the opposite because the algorithm instead rewarded content capable of generating anger, hostility, comments and reactionsan element confirmed by researchers in some internal memos, which read that “misinformation, toxicity and violent content are disproportionately widespread among reshares”.
The Washington Post, working on the same documents, then revealed that every emotional reaction, present among the ways of reacting to a post, including the angry face, was worth five times a like, and that already in 2019 Facebook’s internal researchers had verified how the “Posts capable of eliciting angry reactions via emoji were disproportionately likely to contain misinformation, toxic content and low-quality news”.
What Meta changed on Instagram and Facebook
The reputational damage for Facebook was so profound that the company chose to rename itself Meta. There has been little progress over time regarding the role of the algorithm in amplifying fake news and pushing users towards increasingly extreme political positions. At the beginning of 2025, the company announced the closure of its third-party fact-checking program in the United Statesreplaced by a system of Community Notes written by users on the model of
Meta then recently presented new tools based on artificial intelligence that allow you to tell the platform which contents you want to see in the Feed and in the Reels, a function called “Your Algorithm”. In the release, the company explains that its ranking systems infer interests from what people interact with, from commented posts to videos watched, among thousands of other signals.
In parallel Frances Haugen’s documents became evidentiary material in more than 2,400 lawsuits against Meta, TikTok and YouTubeaccused of designing their apps to addict minors. The first verdict arrived in March 2026 in Los Angeles, in which Meta and Google were sentenced to compensate a twenty-year-old, known as KGM, with a total of 6 million dollars, who attributed years of anxiety, depression and body dysmorphia to the use of Instagram and YouTube since childhood. The more substantial case was instead closed in August with an agreement, in which Meta agreed to pay about 18 billion dollars to 48 American states that from 2023 accused it of making Facebook and Instagram addictive for children, hiding what it knew about the damage to their mental health.
In 2024, Meta introduced accounts for teenagers on Instagram and, following the agreement reached at the end of August 2026, Zuckerberg’s company agreed to introduce stricter constraints and strengthen the protections reserved for minors on Instagram and Facebook in the United States. The new limitations they require a total of 2 hours of use per daya midnight to 6 a.m. lockdown, muted notifications during school hours and more control tools for parents. The experience of using apps also changes. For example on teenagers’ accounts the count of likes and reactions will be hidden and the agreement also introduces a ban, for minors, on filters that reproduce the effects of surgical retouching or make-up that transforms the features.
These new protections, however, do not concern Italy. Here, the security settings that limit contact and inappropriate content remain valid and include private accounts, message restrictions, reminders to exit the app after 60 minutes of daily use and turning off notifications between 10pm and 7am.
Haugen, interviewed by Fortune upon the film’s release, said she not having seen any cultural change that shows the company has turned a corner She said she believes the real solution is to design safer systems from the start, not simply try to clean them up at the end.
