THE WHITE HOUSE says it is attacking violent extremism on the Internet. If so, it is counterattacking at the same time.

Two currents at cross purposes have emerged from President Trump’s administration in recent days, in the wake of yet another mass shooting apparently inspired by white-supremacist rhetoric on the Internet. The first has some potential. Senior officials will meet with executives of top technology companies on Friday to discuss the rise of violent online extremism, a plague Mr. Trump pledged to “do something about” this week. But at the same time, the administration is continuing one of the president’s favorite crusades — against supposed political bias at some of those same companies — in a way that would make it more difficult for them to counter violent extremism.

Another “social media summit” at the White House last month ended with a promise from the president to consider “all regulatory and legislative solutions” to an epidemic of Internet censorship. One problem: All evidence suggests that such an epidemic does not exist. Platforms are merely enforcing their terms of service, which often prohibit hate speech and harassment. Nonetheless, Politico reports that drafts of a proposed executive order are circulating addressing allegations of anti-conservative bias. The president’s actual authority to control online speech is limited, but his ability to rally voters around the idea of a Silicon Valley cabal shutting out dissent is not. Neither is his power to scare companies into letting people break their rules. Firms most fear the sort of “solution” presented by Sen. Josh Hawley (R-Mo.), who has introduced a bill proposing to certify platforms as politically neutral and to remove immunity from lawsuit or prosecution for user-posted content from those who fail that test.

What does this have to do with goal No. 1? Well, “doing something” about white-supremacist incitement on the Internet would require precisely the sort of behavior that Republicans railing against companies for their supposed liberal bias do not want. Sites such as Facebook, Twitter and YouTube would step up efforts to take down racist posts that their standards already prohibit, and perhaps strengthen the standards themselves. They would reduce the recommendation of borderline content likely to lead users down the rabbit hole of radicalization. All these efforts would probably hit conservatives hardest, as in all likelihood would Mr. Trump’s proposed strategy of developing “tools that can detect mass shooters before they strike.” An algorithm that combs the Web for violent manifestos, to the extent it worked at all, would surely capture plenty of strident but nonviolent speech in its sweep.

The White House’s insistence on going after a liberal boogeyman puts in jeopardy any attempt to confront the real threat of violent racism. That puts Americans in jeopardy, too.

Read more: