Joan Didion once wrote that “a good deal about California does not, on its own preferred terms, add up.” Though a land of immigrants, California long had a pronounced nativist streak. A polity with no racial majority, California has a protracted history of racism. A site of unrivaled natural beauty, the state, through its vast corporate farms, accounts for a fifth of the nation’s pesticide use. In California, individualists live comfortably off aqueducts and railroads paid for by the federal government, and pacifists live side by side with the country’s most powerful prison-guard union.
Californians have always been convinced that they know right from wrong, though the “right” and the “wrong” may fluctuate wildly. We used to institutionalize distressed housewives. Now, we let drug addicts roam the streets. What hasn’t changed is our confidence.
Didion observed a “familiar California error, that of selling the future of the place . . . to the highest bidder.” The speculators come and go. The currency of the auction changes. But California is always up for grabs. Today’s robber barons are environmentalists, plaintiffs’ lawyers, Diversity, Equity, and Inclusion hustlers, and a seemingly infinite variety of other activists. They are the inheritors of California’s one enduring tradition: extractive, parasitic governance. California’s present ruling class is a rulemaking class. They don’t really do anything; they earn a living telling others what to do.
Lately, they’ve set their sights on the Internet.
California recently enacted AB 587, a law confusingly titled Content Moderation Requirements for Internet Terms of Service. Nominally a transparency law, it orders social-media platforms to explain how they moderate content. The true aim is clearly to pressure the platforms into moderating content differently. For AB 587 is not about social media content in general; it is, rather, about discrete categories of content that interest progressives. What are you doing, the law asks the platforms, to address “hate speech,” “extremism,” and “misinformation”? AB 587 requires the platforms to submit detailed reports to the government about those subjects specifically. The message is not subtle: the government expects the platforms to crack down on these forms of speech.
What, you might wonder, does a word like “extremism” mean, precisely? AB 587 does not say. It leaves it to the platforms to define the special categories for themselves. Perhaps the lawmakers were not up to the task of giving these slippery notions shape. Or perhaps they saw that their interest lay in ducking it—in letting terms like “misinformation” continue to shift with rapidly evolving elite sensibilities. At all events, the absence of uniformity raises the question of what end the disclosures actually serve. As Internet-law expert Eric Goldman notes, the state seems to want this information “not to benefit any constituency”—after all, it won’t facilitate apples-to-apples comparisons—but as a “raw exercise of power.”
At least AB 587 stops at social media. Another new law, AB 2273, seeks to reshape the entire Internet. Judging from its name—the California Age-Appropriate Design Code—you might suppose that the law seeks to help children safely navigate children’s websites. And that is doubtless what the law’s proponents want. Won’t someone think of the children? is a mighty political mace, one that smashes rational objection like plywood. When her husband, California governor Gavin Newsom, signed AB 2273 into law, Jennifer Siebel Newsom laid the rhetoric on extra thick, exclaiming: “I am terrified of the effects technology addiction and saturation are having on our children and their mental health.” But AB 2273 is about far more than “our children.” The law governs any website run by a commercial entity (small businesses excepted) that faces a reasonable prospect of being accessed by a child—as any website does. And the law will function like a zoning regulation, municipal building code, and workplace safety manual rolled into one.
Under AB 2273, companies will have to produce a “data protection impact assessment” for every “product, service, or feature” on their websites. These assessments are redolent of the environmental-impact statements that so effectively hinder building projects in the physical world. What is the “purpose” of the product, service, or feature in question? Could it attract children to the website? Will it collect or process information about them? Could it “expos[e]” them to “potentially harmful” content? These are just a few of the many questions that each assessment will need to answer.
Yes, all but the most child-friendly websites could expose a child to potentially harmful content. The front page of a newspaper could, too. “Maybe” will therefore be the appropriate answer to many of the questions that AB 2273 poses—a response that pressure groups, the state, and the media will happily treat as an admission of guilt.
In theory, the assessments will not be public documents. In practice, however, the state attorney general will be able to obtain them upon request. Sacramento talks piously of data privacy; it does not take data privacy seriously itself. Last year, the attorney general’s office inadvertently exposed the personal information of hundreds of thousands of the state’s gun owners. We can expect disfavored websites to suffer similar rough treatment. And we can only shrug at how, in California, government-mandated data-protection assessments will be subjected to the government’s own lax standards of data protection. The Californian model of statecraft is a flywheel of hypocrisy.
When not churning out the assessments, lawyers will be busy determining whether websites offer children a “high level of privacy” by default. They will be composing plans “to mitigate or eliminate” the risk of “material detriment to children” that might arise from websites’ “data management practices.” They will be guessing at whether websites use “dark patterns” (a term that AB 2273 only vaguely defines) in a way that could diminish children’s “well-being.” They will be trying to draft terms of service in “language suited to the age of children likely to access” a given website. (Good luck, CoComelon.) They will be monitoring whether companies “consider the best interests of children” and “prioritize the privacy, safety, and well-being of children over commercial interests” when creating websites. AB 2273 is sure to benefit at least some children: attorneys will be topping up their kids’ college funds.
Companies will have to pay the attorney general thousands of dollars in penalties per child “affected” by a violation of the code. That money, in turn, is slated to fund further enforcement. Meantime, the law will establish a Children’s Protection Working Group, whose unmistakable task will be to recommend augmenting and strengthening the scheme of regulation.
Does the Internet harm children? The evidence, unsurprisingly, is mixed (and compared with what?). But the sensible approach, if we’re worried, is for academics to keep investigating the matter; for families and schools to teach children responsible Internet habits; and for nonprofits and the government to highlight the many online controls available to parents.
California’s approach, by contrast, is one of overreaction and centralized control. In one prominent tech commentator’s words, AB 2273 “insists that there’s a problem—without the evidence to back it up—and then demands an impossible solution that wouldn’t actually fix the problem if it were a problem.” The law will make website design a plodding and uncertain exercise in risk aversion. It will prod firms to sanitize and self-censor, lest they have to report “harming” children. Perhaps worst of all, the law will push websites toward verifying the age, and likely the identity, of users. That would be not just a nuisance but a disaster for privacy. Imagine having to submit your identification to almost every website you visit. (Don’t worry, say some enterprising firms: they’ll use facial recognition instead!)
AB 587 and AB 2273 are just the beginning. California lawmakers are also regulating how social-media users report cyberbullies, as well as when and how other states can access data related to abortions. They’re considering a child online “addiction” bill that would expose platforms to open-ended liability for the crime of offering appealing products. (A child “preoccupied” with, or having “difficulty” not using, social media is an addict, as defined by the bill.) And the California Privacy Protection Agency is busily emitting directives for the state’s strict data-privacy law. The common thread in these laws, bills, and rules is that they’re all national in scope, to varying degrees. They attempt to govern the Internet for everyone.
Californians mined the Sierra Nevada and plowed the Central Valley. They built Paramount Pictures and the Douglas Aircraft Company. When there were no wilds left to explore, they dreamed up a new one—the Internet. But a state repeatedly made rich by enterprise, opportunity, and risk is now resentful, cautious, and hidebound. The digital frontier, we’re told—by Californians, of all people—is closed.
Even now, a line runs from Sutter’s Mill to Silicon Valley. But if California is not careful, its latest gold rush will be its last.
Photo: 2d illustrations and photos/iStock