When Arturo Béjar took the stand in an Oakland courtroom for a second consecutive day, he wasn't a hostile witness dragged in by plaintiffs' attorneys. He was someone who spent years inside Meta trying to fix the exact problems he's now describing under oath — and who left, in part, because the fixes never came.

Béjar, who served as an engineering director at Meta and previously testified before a U.S. Senate subcommittee on child safety, told jurors that Instagram operated under what he called a "don't ask, don't tell" policy toward users under 13. The platform's own terms of service bar children below that age — a requirement rooted in the Children's Online Privacy Protection Act, or COPPA — but Béjar's testimony describes a corporate culture that preferred not to know how many minors were on the platform, because knowing would have created legal and regulatory exposure.

The distinction matters enormously. There is a significant legal and moral difference between a platform that cannot detect underage users despite genuine effort, and one that designs its systems to remain deliberately blind to their presence. Béjar's account places Meta squarely in the second category. If the jury believes him — and he is a credentialed, firsthand witness — the implications for Meta's liability calculus are severe.

The trial unfolding in Oakland is one of the most consequential legal proceedings involving a major social media company in the industry's history. It consolidates claims from families who allege that Instagram's design — its recommendation algorithms, notification systems, and engagement architecture — caused measurable psychological harm to minors, including depression, eating disorders, and in some cases self-harm. Meta has maintained that its platforms are safe when used as intended and that responsibility for underage access lies partly with parents.

Béjar's testimony directly undercuts that defense. According to his account, Meta's product teams were not primarily optimizing for user wellbeing — they were optimizing for time-on-platform and interaction frequency, metrics that feed advertising revenue. He testified that safety concerns he raised internally were consistently subordinated to engagement goals. This is not a case of an engineer speculating about corporate culture from a distance; Béjar was in the rooms where product decisions were made.

Meta's internal research has already become part of the public record through prior Congressional proceedings and document productions in related litigation. That research, produced by Meta's own teams, found that Instagram use was associated with negative body image outcomes and mental health effects in teenage girls — findings the company did not publicize. Béjar's live testimony gives those documents a human narrator: someone who saw the data, raised the alarm, and watched leadership set it aside.

The "don't ask, don't tell" framing is legally precise in a way that will not be lost on the jury. It implies not ignorance but willful ignorance — a deliberate policy architecture designed to preserve plausible deniability. COPPA imposes significant obligations on platforms that knowingly collect data from children under 13. By structuring its systems to avoid confirming the age of young users, Meta may have been threading a legal needle. Whether that threading constitutes a violation, a fraud on regulators, or simply aggressive compliance lawyering is exactly what this trial is designed to test.

Meta has not disputed that Béjar worked there or that he testified before Congress. The company's legal strategy appears to center on contesting causation — arguing that even if minors used Instagram and even if they experienced mental health difficulties, the platform cannot be proven to have caused those difficulties. That argument has traction in some academic literature, which remains genuinely contested. But it becomes much harder to sustain when a former senior engineering director is describing, from firsthand knowledge, a company that saw risk signals and chose revenue over response.

The Oakland trial represents a reckoning that federal regulators have so far failed to deliver. Congress has held hearing after hearing, senators have waved printed slides, and the Federal Trade Commission has issued reports. None of it has structurally changed how Instagram is built. A jury verdict — particularly one that attaches significant damages — could do what legislation has not: force a redesign driven by liability rather than lobbying.