Meta 'knew and did nothing,' whistleblower reveals in child safety trial
Meta is currently undergoing a landmark trial in Oakland, California, litigated by four US states — Colorado, California, New Jersey and Kentucky. The case could permanently change how popular social media...
Meta is currently undergoing a landmark trial in Oakland, California, litigated by four US states — Colorado, California, New Jersey and Kentucky. The case could permanently change how popular social media platforms such as Facebook and Instagram are designed.
The states accuse the social media giant of intentionally designing its platforms to addict younger users and hiding what it knew about the harmful effects on their mental health, claims Meta has consistently denied.
The trial, which started on Tuesday, has been called the largest consumer protection lawsuit in American history by Kentucky's attorney general, Russell Coleman.
Testimony in the trial's first three days has already produced striking claims about what Meta knew and when.
Whistleblower Arturo Béjar, a former Meta engineering director, testified that the company's public Community Standards Enforcement Reports — which measure how often users are shown content that violates Meta's policies — could be understating the real risk by as much as 100 times, based on internal user surveys that asked people directly about their negative experiences on the platforms.
'Hook, harvest, hide'
California's deputy attorney general, Megan O'Neill, told jurors in her opening statement that Meta's business model could be summed up in four words: "Hook the users. Hold them for as long as they can. Harvest their data. Hide the truth from the public when making public statements."
She alleged that Meta hid the truth about the real risks of its platforms for younger users in an attempt to hook them and harvest their data, and accused the company of actively intervening in its own mental health research to further obscure those risks.
Meta's lawyer, Paul Schmidt, argued in response that the tech giant has a track record of sharing information about social media harms, and that enforcing content and age-limit policies is complicated. Verifying a user's age through documents such as government identification or credit cards, he said, poses its own problems, since not every user has access to either.
Schmidt also said Meta's employees are working on improving the platforms' impact on user wellbeing, and urged the jury to keep an open mind.
Prosecutors also argued that Meta was aware large numbers of children under 13 were active on Instagram despite the platform's rules.
Béjar testified that his own research had uncovered tens of thousands of under-13 users, while Meta's own technology was capable of flagging potentially millions more.
Rather than pushing to remove them, he said, the company's approach was effectively "don't ask, don't tell."
Béjar told jurors that despite repeated warnings, Meta continued to treat safety as an afterthought — including during the rollout of major features such as Reels, where he said "safety was not a consideration in how it was initially deployed."
He described emailing Chris Cox, Meta's chief product officer, to flag what he saw as a stark gap between the harm users were reporting and what the company was prioritising and telling the public.
Béjar has previously said that in an earlier conversation, Cox was the first Meta executive who could answer, off the top of his head, what percentage of users were experiencing this kind of harm — a moment Béjar described as "heartbreaking," because it meant Meta's leadership already knew and had chosen not to act.
Meta primarily measures harm on its platforms through the prevalence of content that violates its own policies.
Béjar rejected this metric outright. Citing the example of a 13-year-old sharing distress and asking for help, he told the court: "Harm is what people experience as harm. Prevalence is a very narrow measure of a very specific thing."
He also said Instagram had shifted from a platform users actively chose to use, to one that increasingly used them — largely because its feed evolved from showing mostly people users followed to showing mostly strangers, a change he said introduced substantial new risks.
The outcome of the trial could have ripple effects well beyond Meta. Because the case rests on the argument that a platform's interface and design choices — not just individual pieces of content — can create legal liability, a ruling against Meta could shape how rivals such as Snapchat, TikTok and YouTube approach the same questions.
