Why Conservatives Are Wrong to Support the Kids Online Safety Act

A fter languishing for years, the Kids Online Safety Act (KOSA) finally cleared the Senate in July. Yesterday, the House Energy & Commerce Committee (ECC) approved the bill, readying it for a floor vote in the lower chamber.

Conservatives committed to free speech must once again explain that KOSA vests censorship authorities in an already too-censorial federal bureaucracy. The bill has undergone myriad revisions as legislators have attempted to edit away its clear deficiencies. But these deficiencies are fundamental to its structure — excising them would also level its foundation.

KOSA’s loudest supporters include conservatives rightly eager to shield children from the internet’s nastier aspects. In their well-meaning haste, though, these conservatives have papered over the bill’s flaws. Proponents argue that something must be done to protect kids online, so we ought not look too closely at the mechanism by which Congress proposes to do so. History teaches that the all-too-human urge to do something ­— anything — to ameliorate perceived problems often produces frustratingly permanent regulatory structures. These often not only fail to achieve their stated ends but threaten liberty, sometimes in unforeseen ways. Prudence dictates that remedies be considered carefully and tailored to the problem at hand.

Even some of KOSA’s friends on the committee voiced concern. “We might be back in here sooner [or] later doing hearings on how the FTC [Federal Trade Commission] or social-media companies have conducted very serious content throttling and censoring in order to avoid the liability that this new law would impose,” argued Representative Dan Crenshaw (R., Texas) at yesterday’s markup, although he said he would still vote for KOSA.

The primary danger of KOSA lives within its “duty of care,” which requires online platforms to shield children from a range of broad categories of harmful content. Of course, in a perfect world, free from tradeoffs, children would never face such problems. But lawmakers must consider what flows naturally from a regulatory regime requiring online platforms to suppress certain types of speech.

Conservatives who support KOSA protest that the listed disorders have clear definitions. This is true to an extent — though of limited usefulness. What’s unclear is what kinds of content regulators will deem to contribute to such disorders. And as Representative Jay Obernolte (R., Calif.) said, the bill provides little clarity regarding what its duty of care would require of platforms (it employs the notoriously malleable reasonableness standard). Precisely how far the would require platforms to go to keep purportedly dangerous content from children remains entirely unclear. As Crenshaw noted wryly, “Doesn’t all political speech induce some kind of emotional distress for those who disagree with it?”

Such discretion could easily give way to censorship. Ideologically motivated enforcers would likely stretch the statutory text to the breaking point, and platforms would likely remove large amounts of content to avoid even a whiff of liability. Recent history ought to give pause to conservatives ready to authorize bureaucrats to censor online speech, even with respect to children.

The ECC amended the bill directly before passing it out of committee, softening many of its sharpest and most unconstitutional edges. These changes advanced in spite of the objections of many representatives, who said they robbed the bill of its potency. It seems likely that future negotiations will pivot KOSA back towards its original, more pernicious form, which more closely resembles the version passed by the Senate. Neither the Senate nor the more aggressive members of the House will easily allow the wings of their pet project to be clipped. Even the amendment’s supporters assured their more hardline and skeptical colleagues that further alterations — which, presumably, would realign the House’s version with the Senate’s — would follow the bill’s ascension from committee. And it should be noted that even with the ECC’s amendments, KOSA’s constitutional problems remain severe.

The liability KOSA imposes would mean stifling users’ speech — including much speech protected by the First Amendment. In some cases, children (and society generally) would not miss the censored speech, and much of this censorship would likely fall in court anyway. But especially in its Senate form, KOSA would still lend a congressional imprimatur to what most Americans would recognize as rank censorship.

The rejoinders of KOSA’s conservative advocates give little comfort. Confronted with the bill’s potential for censorship and abuse, many conservatives have simply denied its plain text. In fact, in their more open moments, they have confirmed the fact that KOSA provides ideologically driven enforcers with a censorial cudgel. Senator Marsha Blackburn (R., Tenn.), one of KOSA’s most enthusiastic fans, says the bill could prevent children from “being indoctrinated.” According to a recent pro-KOSA memo from the Heritage Foundation, “A Republican administration could fill the [Kids Online Safety Council, which KOSA would create] with representatives who share pro-life values.”

This admits — quite bald-facedly — that the ideological makeup of the agencies enforcing KOSA will matter a great deal. Indeed, last year, Heritage explicitly advocated deploying KOSA to “[keep] trans content away from children.”

Whatever one thinks of hot-button social issues, conservatives should not naively presume that the left would not manipulate KOSA’s provisions any less creatively or enthusiastically than they would. If a Trump administration could suppress transgender content, a Harris administration could suppress non-leftist views on guns, sex and gender, public health, and more.

The panic over children’s online safety is not entirely unfounded. The desire to protect kids is noble. But conservatives should know better than to trust the vigorously disputed evidence of a small cadre of experts and, consequently, to rush to support measures that threaten basic civil liberties. “Trusting the science” — particularly when “the science” has been hurried and sloppy — rarely promotes either liberty or security. An epistemologically prudent view favors working at a measured pace toward reforms that are consistent with our constitutional guarantees.

In policy-making, tradeoffs are inevitable. However, the tradeoffs that KOSA requires are intolerable to a free society. Other means of protecting children online must be pursued instead.