David French of National Review Online highlights a key problem related to social media censorship.

… Silicon Valley has lost the public trust of tens of millions of Americans. They know the rules are malleable. They know double standards apply. And they know that the campus-censorship culture is being imported online.

On campus (and increasingly also on social media), far-right anger and attacks are deemed dangerous and abusive. Far-left anger and attacks are instead deemed expressions of righteous outrage. Maliciously racialized far-right language is seen as evidence of white supremacy and white nationalism. Maliciously racialized far-left language is perceived as an attack on the privileged and powerful.

The result is that hate-speech policies exist not as easily interpreted, uniformly applied rules that provide all users fair notice of the conditions for using the platforms, but rather as subjectively interpreted, selectively applied weapons to wield on behalf of favored ideas and individuals. The result is that some people are more exposed to abuse than others because those people are deemed less worthy of protection.

Twitter will move to protect a U.S. congresswoman like Ilhan Omar — a highly visible public figure with a huge platform — from attacks on her faith, but it will not lift a finger to protect an unknown elderly woman from becoming an object of hate and derision on the basis of her age, race, or faith. How does this make any rational sense?