Stripped of attorney-client privilege by a rare court order last month, Meta is now confronting the release of internal documents that allege a systematic cover-up of platform harms.
Unsealed filings from the multidistrict litigation (MDL) suing the tech giant reveal that executives terminated “Project Mercury”, a 2020 study confirming Facebook’s causal link to depression, to shield the company from liability.
The disclosures further expose a “17-strike” threshold for sex traffickers and text messages where CEO Mark Zuckerberg allegedly rejected safety funding to prioritize “building the metaverse,” directly contradicting the company’s public defense of its safety record.
Project Mercury: The Buried Evidence of Causality
Buried within the unredacted court filings is the detailed history of “Project Mercury,” a research initiative launched in 2020 to definitively answer whether Facebook harms user mental health. Unlike previous observational studies, this project employed a rigorous experimental design in collaboration with survey firm Nielsen.
Scientists instructed a test group of users to deactivate their Facebook accounts for one week, measuring their mental state before and after the hiatus.
Results from the experiment were unambiguous. Internal records show that the study identified a direct link between platform use and negative psychological outcomes. Participants who stepped away from the platform reported measurable decreases in depression, anxiety, and loneliness.
According to the court document, the study found that “[p]eople who stopped using Facebook for a week reported lower feelings of depression, anxiety, loneliness, and social comparison.”
One Meta employee reportedly warned about the negative implications about withholding the study results, saying “if the results are bad and we don’t publish and they leak, is it going to look like tobacco companies doing research and knowing cigs were bad and then keeping that info to themselves?”
Validating the data’s significance, an unnamed staff researcher noted in an internal chat log that “The Nielsen study does show causal impact on social comparison.”
A senior data scientist at Meta who reportedly “also holds a PhD in neuroscience […] and taught a university course on addiction” warned:
“It seems clear from what’s presented here that some of our users are addicted to our products. And I worry that driving sessions incentivizes us to make our product more addictive, without providing much more value. How to keep someone returning over and over to the same behavior each day?
Intermittent rewards are most effective (think slot machines) reinforcing behaviors that become especially hard to extinguish—even when they provide little reward, or cease providing reward at all.”
Despite the scientific value of these findings, the filings allege that Meta executives moved quickly to suppress them. Rather than publishing the data or using it to re-engineer the product, leadership reportedly terminated the project.
Case 4 22-md-03047-YGR court document
Justification for the shutdown did not cite technical errors but rather reputational risk. Documents indicate that executives argued the results were tainted by the “existing media narrative” surrounding the company, a rationale that effectively prioritized public relations over user safety.
Frustration among the research team was palpable. Drawing a parallel to historical corporate malfeasance, one employee lamented in a chat message that the company was “doing research and knowing cigs were bad and then keeping that info to themselves.”
Meta has forcefully rejected this characterization of events. In a statement responding to the unsealed filings, spokesman Andy Stone argued that the study was halted due to “flawed methodology” rather than its conclusions. He further stated that “We strongly disagree with these allegations, which rely on cherry-picked quotes and misinformed opinions.”
Yet, the plaintiffs argue that the termination of Project Mercury was not an isolated incident but part of a broader strategy to avoid creating “causal” evidence that could be used in litigation or regulatory hearings.
The ’17-Strike’ Rule: A Pattern of Calculated Negligence
Beyond the suppression of research, the amended master complaint details specific operational policies that appear to contradict Meta’s public stance on child safety. Perhaps the most significant revelation is the existence of a “17-strike” policy for handling sex trafficking accounts.
Testimony from former safety staff indicates that the platform’s enforcement threshold was set exceptionally high. Under this rule, a user reportedly had to be flagged 17 times for attempting to traffic people for sex before their account would be removed.
Internal documents described this policy as “a very, very, very high strike threshold,” a setting that allowed predators to remain active on the platform despite repeated violations.
Listing further grievances, the filing details systemic failures:
“Meta intentionally designed its youth safety features to be ineffective and rarely used, and blocked testing of safety features that it feared might be harmful to growth.”
“Meta recognized that optimizing its products to increase teen engagement resulted in serving them more harmful content, but did so anyway.”
Far from a passive oversight, the filings suggest that the ineffectiveness of safety tools was a deliberate design choice. Plaintiffs allege that features intended to protect youth were engineered to be difficult to locate or use, ensuring they did not interfere with engagement metrics.
Exposing a rift in the C-suite, the documents highlight a direct conflict between the company’s policy leadership and its CEO regarding the funding of safety initiatives.
When Nick Clegg, then-head of global public policy, requested additional budget to bolster child safety measures, Mark Zuckerberg reportedly denied the request via text message.
Justifying the refusal, Zuckerberg wrote that he would not prioritize those requests “when I have a number of other areas I’m more focused on like building the metaverse.”
Occurring in 2021, during the company’s aggressive rebranding from Facebook to Meta, this exchange provides documentary evidence supporting the narrative that the company diverted critical resources away from platform safety to fund its hardware ambitions.
Piercing the Veil: The Crime-Fraud Exception and Global Fallout
These disclosures were precipitated by a significant legal defeat for Meta in October. As covered in previous reporting, Judge Yvonne Williams of the D.C. Superior Court invoked the rare “crime-fraud exception” to pierce the company’s attorney-client privilege.
Usually, communications between a company and its legal counsel are protected from discovery. However, the court found that this protection did not apply in this instance because “Meta’s counsel offered such legal advice to specifically limit Meta’s potential liability.”
Effectively stripping Meta of its ability to shield these internal documents, the ruling validates long-standing accusations that the company’s legal team actively managed research to avoid creating a “paper trail” of liability.
Public scrutiny of these internal records arrives at a moment of intense global pressure for the social media giant.
In Australia, the company has begun mass account deactivations to comply with a strict new federal ban on social media for users under 16. That “digital eviction” represents the first time a Western democracy has mandated the total exclusion of minors from these platforms.
Simultaneously, European regulators are escalating their own enforcement actions. Spanish Prime Minister Pedro Sánchez recently ordered a parliamentary investigation into the company, labeling the current social media landscape a “failed state” that requires a fundamental “refounding.”
With the landmark child safety laws recently enacted in California and the ongoing multidistrict litigation gaining new evidentiary ammunition, the era of industry self-regulation appears to be definitively over.
