The most damaging witnesses in corporate accountability trials are almost never regulators or outside experts. They are the people who sat in the rooms. On the opening day of a trial brought by 29 states against Meta Platforms, the lead witness was Arturo Bejar — a former Meta engineering director who left the company and then, unusually for Silicon Valley, kept talking.

Bejar's testimony went to the core claim of the states' case: that Meta's leadership, including CEO Mark Zuckerberg, was presented with evidence of harm to young users and made a deliberate choice to prioritise platform growth. That is a different allegation from negligence. Negligence means you didn't know. This says they knew, weighed it, and decided engagement metrics mattered more.

The trial, which opened in a California court and is expected to run six weeks, consolidates claims from California, Colorado, Kentucky, New Jersey, and 25 additional states. The states allege Meta designed Instagram and Facebook to psychologically hook young users, and that this design contributed to anxiety, depression, and in some cases suicide among adolescents. They further allege the company violated federal child privacy law by collecting and using personal data from children under 13 without lawful parental consent, while publicly misrepresenting the platforms' safety record to parents and the public.

Zuckerberg is expected to testify. That is not a routine development. CEOs of major technology companies are rarely compelled to take a stand in state-level civil proceedings; when they are, it signals either that the documentary record is close enough to require his explanation, or that the states' legal position is strong enough that a settlement was not reached. Both possibilities are uncomfortable for Meta.

The internal documents underlying the states' case have been referenced in prior congressional testimony and regulatory filings, including material that surfaced during earlier whistleblower proceedings. What those documents showed, in summary, was a company with its own internal safety research flagging harm signals and a product organisation that was structured to make growth the overriding variable. Meta has disputed that characterisation, arguing its safety investments are substantial and genuine.

The federal child privacy dimension of the case rests on the Children's Online Privacy Protection Act, which prohibits platforms from collecting personal data from users under 13 without verifiable parental consent. The states allege Meta collected that data anyway, through a combination of inadequate age verification and deliberate design choices that made underage use easy and invisible. This is not a novel allegation — the Federal Trade Commission has previously taken action against Meta on related grounds — but the scale of the state coalition here and the trial setting give it a new level of legal exposure.

What makes the Bejar testimony particularly significant is its source. This is not an outside researcher publishing an academic paper, or a journalist reconstructing events from leaked documents. This is someone who held a senior technical role inside the company, attended the relevant meetings, and is now testifying under oath that what he observed was a leadership aware of the harm and uninterested in letting it slow the product. Internal witnesses do not usually emerge in American tech litigation. When they do, juries and judges notice.

Meta's legal team will work to establish that Bejar's account is selective, that his departure created a grievance, and that the company's subsequent safety investments demonstrate good faith. Those are legitimate trial arguments. They are also arguments that require Meta to explain, in a courtroom, why so many of its own internal safety recommendations were not acted upon when Bejar and others raised them. That explanation, when it comes, will matter more than the press statements already in circulation.