In this photo illustration a LinkedIn logo is seen displayed on a smartphone screen with a computer keyboard in the background in Athens, Greece on Jan. 10, 2022. Photo: by Nikolas Kokovlis/NurPhoto/Reuters

I did not write a slur. I did not threaten anyone. I drafted one sentence to a friend on LinkedIn.

He had criticized me for attacking JD Vance’s refusal to condemn or distance himself from Tucker Carlson. My friend called that criticism political. I found the charge offensive for two reasons. I do not treat antisemitism as a partisan issue. And I have a long public record of criticizing antisemitic Democrats (as well as Republicans).

So, I wrote back: “Definitely not [political]. You’ve seen me endlessly go after the antisemitic Dem politicians.”

I clicked the “post” button but was surprised to see that the reply did not go up.

Instead, a LinkedIn popup warning appeared. Linked In claims it belongs to the company’s “kindness reminder” system: software that flags “language that has been reported in the past,” then asks, in effect, whether you are sure you want to post it.

Three things about that pop-up should bother anyone who uses a professional network as if it were still a place for adults.

First, the sentence was not offensive. It named a form of hatred and said I oppose it on all sides.

Second, LinkedIn is not my mother. I have a reputable profile on a site that sells itself as a workplace. This is still America. A machine should not scold me, or anyone, for a civil reply.

Third, and perhaps worst of all: Why has the word “antisemitic” been reported so often that the filter now treats it as a problem? At this rate, will “Jew” be the next problematic word we are asked to reconsider using?

Let’s take a step back for a moment and look at LinkedIn’s record. After Oct. 7, 2023, Australia’s Online Hate Prevention Institute began watching the platform because staff saw a surge of Jew-hatred there. Across 10 sites, LinkedIn had the best removal rate (36%) for reported antisemitic posts.

But most of the hate stayed up. Researchers later re-reported posts, including “Hitler was right” material, and were told some still complied with LinkedIn’s rules.

LinkedIn’s policies do not even name antisemitism. In testimony to Australia’s Royal Commission, LinkedIn’s James Corrigan said internal rules treat “Zionist” and “Israel” as stand-ins for attacking Jews when used that way, but the company had not published that. He confirmed another design choice: a reported post disappears from the reporter’s feed and stays visible to everyone else until a later decision. He said LinkedIn cannot quantify how much antisemitic content it hosts

Outrageously, Jewish users who try to fight that content have been punished for the effort. Manhattan CPA Marc Strohl said he reported posts such as “Death to Israel” and claims that Jews were animals who “need to be extinct.” He said LinkedIn often did nothing, then banned him. He sued. The account was later restored after a settlement, amid claims of selective enforcement against Jewish and pro-Israel users.

That is the machine that tapped me. It trains on language members have reported. But reports are not neutral weather. They are often insidious, well-funded, and massively widespread tactics.

The Center for Countering Digital Hate found that when researchers reported antisemitic posts in large Facebook groups, admins banned the reporters from the original group and from affiliated groups that shared moderators. The report button was used against the people reporting Jew-hatred.

The Jerusalem Post has described Telegram networks that assign Jewish Instagram accounts as homework: comment, flood DMs, report the profile, screenshot the work. Melanie Shiraz wrote that death-wish comments against her family stay up while her own inoffensive posts vanish. “Mass reporting campaigns work,” she wrote. “The algorithm responds to volume, not validity.”

Jews are a miniscule population in a digital ocean. Coordinated reports beat scattered good faith reports every time.

The Online Hate Prevention Institute put the method in one sentence: Efforts to silence concerns about antisemitism create space for antisemitism to grow. AJC found that 65 percent of American Jews who saw antisemitism online did not report it, most often because they did not think the platform would act. ADL’s scorecards have shown major platforms routinely failing to remove antisemitic hate reported through ordinary channels. Instagram testing in 2026 left 93 percent of reported extremist and hateful content in place.

I cannot prove a foreign service wrote LinkedIn’s popup. But I can show the pattern: Flood the report button. Recode “antisemitism” as a political smear. Train the model until the word itself looks dirty. Then decent people hesitate, and the hate stays.

LinkedIn may not have called me a bigot. But it treated the key vocabulary piece of moral anti-Jewish description as a risk factor. If a professional network cannot tolerate the word “antisemitic” in a sentence that condemns antisemitic politicians, the antisemites have already won a fight they never had to win in public.

Unless concrete steps are taken by social media companies to reject the “might is always right” framework in the reports and complaints arena, all minority groups, not just Jews, face terrifying prospects. Without action, how long will it be until you can’t say “Jew” on social media at all?

Jeffrey Lax is a professor of law at the City University of New York, founder of S.A.F.E. Campus, and a frequent legal and political commentator on Fox News, Newsmax, and other major media outlets. Follow him on X at CUNY_Prof.