It is hard to see how society will be worse now that Facebook has banned Louis Farrakhan, Alex Jones and assorted other hateful extremists from its services. Will American political discourse really be poorer for the absence of Farrakhan’s ranting about “the Satanic Jew” or Jones’s warnings about the Pentagon’s “gay bomb”? The question answers itself.

But another question doesn’t: Who will they ban next? Do we really want to give a handful of tech companies effective control over what Americans may know, and from whom they may get their information? There is good reason that courts have largely protected speech, full stop, not merely speech that doesn’t offend anyone. Deciding what constitutes offense almost inevitably ends up serving the ideological biases, and the interests, of the class of people doing the deciding.

Though, to be sure, we already get our information mostly from a relatively small number of people, who collectively decide what the nation as a whole will learn about. We call them news organizations and, until the Internet came around, that was the only way most people had to find out what was happening. Those organizations have historically made judgments about offensiveness in much the way that Facebook, Twitter and social-media platforms will probably make them: a combination of personal values, professional judgment and fear of making customers angry enough to cancel their subscription.

But that assumption might not completely satisfy the public — or even the social-media giants.

The public might worry because social media is much more consolidated than the old media industry ever was. Yes, the average U.S. citizen in 1970 had access to three television networks and one daily newspaper. But there were also countless magazines available by mail, which acted as a substantial check on local censorship. Some local stories might have been overlooked, but at the national level, there was robust competition.

Social media, by contrast, tends to consolidate down to a handful of players at the national — or even the global — level, simply because everyone wants to be in the same social media networks as their friends. Moreover, those networks are largely run out of a single small area that doesn’t look much like the rest of the country — economically, demographically or ideologically. Little wonder that many people in “flyover country” distrust the decisions those companies are making about what is acceptable and what is verboten.

But it’s worth noting that essentially the same thing is happening with old media as the market consolidates into a handful of outlets in just a few cities. To the extent that people worry about censorship, their anxiety is better focused not on Facebook or Twitter, but on the economic and technological forces denuding the media landscape of most of its diversity.

The comparison with a consolidated old media might take the heat off social media companies, but they’re going to be reluctant to use it, because that isn’t a comparison they want to invite. Facebook and other companies may want the freedom to exercise some editorial judgment over what can run on their platforms and what cannot — and they may feel they must lest customers flee from toxic content. But for the old media, that freedom came with great responsibility: Careless editorial judgment could mean being sued for libel. And that’s a headache social media companies definitely don’t want.

The platforms are immune from such suits under Section 230 of the Communications Decency Act of 1996. The law treats them as a neutral pass-through — something like a community bulletin board — and doesn’t hold them responsible for what users post there. That is eminently practical given the sheer volume of material the platforms have to deal with. But it creates a certain tension when a company such as Facebook argues that it has every right to kick off people who say things it considers abhorrent.

Facebook is acting more and more like a media company, with a media company’s editorial oversight (not to mention an increasing share of the industry’s ad revenue). If Facebook is going to behave like a media provider, picking and choosing what viewpoints to represent, then it’s hard to argue that the company should still have immunity from the legal constraints that old-media organizations live with.

At the moment, social media platforms have nothing to fear, for they’re still protected by Section 230. But the more they come to dominate the consumption of information, and the more they use their power to shape it, the more likely it is that the law will change.

Read more: