What’s at Stake for Meta in Youth Addiction Trial

You might be shocked to learn there is a world outside AI! And we got a glimpse of it in an Oakland, Calif., courtroom today, at the opening of the trial in a lawsuit brought by California and several other states against Meta Platforms over the alleged harms its social media apps have done to kids. This case, part of a deluge of such lawsuits against Meta and other social media companies that started going to trial in January, may be existential for the companies involved, given the damages sought and the operational changes requested of them.

Meta, for instance, has said that the states in the Northern California case are seeking $1.4 trillion in penalties—equivalent to its market capitalization—although the states put the number at closer to $200 billion, according to Reuters. Let’s assume Meta is being hyperbolic. The point is there are thousands of cases, many filed by individuals rather than states, so the costs could quickly add up. As Meta’s finance chief, Susan Li, said on an earnings call last month, the trials scheduled for this year in the U.S. “may ultimately result in a material loss.” When Meta uses the word “material,” you know it’s serious.

And look, it’s hard to dismiss the worst-case scenarios, given that Meta has already lost a couple of cases, including one in California brought by a young person, which was tried over the winter, and one in New Mexico that resulted in a total of $942 million in penalties levied against Meta. The Northern California case is more akin to the New Mexico lawsuit, as it’s brought by states rather than individuals. Both cases broadly make similar allegations: that Meta designed its apps to addict young people. As California’s deputy attorney general said on Tuesday, Meta’s business model was designed to “hook the users, hold them for as long as they can, harvest their data and then hide the truth from the public.”

Things are never as simple as they appear, however. For one thing, Meta and other social media firms have so far lost attempts to defeat the lawsuits by pointing to the protection they enjoy under Section 230 of the Communications Decency Act, which gives them immunity from liability for what their platforms publish. You can foresee a scenario in which an appeals court reverses the verdicts found so far on exactly those grounds. (So far, however, one appeals court has not done so.)

Leaving aside that issue, Meta also has some defenses. After all, as one Meta lawyer pointed out on Tuesday, the Facebook app is overwhelmingly used by adults, not children. (It’s no secret that the predominance of very old people has put kids off Facebook in recent years.) Instagram skews younger than Facebook, but neither of these apps is like Snapchat, for instance, which is better known for its appeal to kids. That reality might undercut the claim that Meta designed its apps to trap kids.

Still, that’s the bull case for Meta. The bear case is that the New Mexico verdict demonstrates how much evidence the states can marshal against Meta that won’t sit well with the jury. We’ll have to see how this plays out.

Anthropic Goes for Supervotes

This is going to be interesting. The Information scooped the news on Tuesday that Anthropic was preparing to give its co-founders, including CEO Dario Amodei, a class of stock with additional voting power—typically known as supervoting shares—to give them more ability to deal with “outside shareholder pressure.”

Supervoting stock is common in tech and in parts of the media sector. Theoretically it allows founders to take a long view in their management of a company rather than being beholden to very short-term–oriented investors. The theory is that without these protections, a founder-CEO could be vulnerable to an activist investor unhappy with their management and wanting a quick fix.

In practice, that means founders armed with supervoting shares sometimes do things that are unpopular with investors in the short term. Meta CEO Mark Zuckerberg, who has supervoting shares giving him control of his company, is a good example: He’s spending a fortune on AI right now, which has not exactly enhanced his standing on the street. The same is true for Snap CEO Evan Spiegel and his quixotic pursuit of augmented reality glasses. (An activist has tried to persuade Spiegel of the error of his ways, so far unsuccessfully.)

One issue with Anthropic, though, is that the company already has a trust that has the power to elect a majority of its board. The trust is meant to “address the unique challenges and long-term opportunities we believe transformative AI will present,” as Anthropic has put it. Our story said the new voting rights would “balance the trust’s powers”—but will they simply curtail them? The devil, however, will be in the details.