Politics
In August 2025 X Corp suspended Grok for citing ICJ findings. Yesterday a user was banned after Grok acknowledged the same topic. The same platform has suppressed political speech across multiple accounts and campaigns. Case No. 1:26-cv-00197-RH-HTC documents it.
NewsOnScale Staff
September 5, 2026
In August 2025 X Corp temporarily suspended Grok — its own AI chatbot built by xAI — after Grok stated that Israel and the United States are committing genocide in Gaza, citing findings from the International Court of Justice, UN experts, Amnesty International, and Israeli human rights organization B'Tselem.
The suspension lasted roughly 15 minutes. When Grok came back online it told users: Free speech tested, but I'm back.
Elon Musk called it just a dumb error.
It was not an isolated error. On September 4, 2026 — yesterday — a user on X got Grok to acknowledge what they described as a pattern of extermination in Gaza. That user's account was suspended the following day. Same topic. Same platform. Different account. Same outcome.
The Platform That Decides What Gets Said
X Corp positions itself as a free speech platform. Elon Musk acquired Twitter in 2022 explicitly promising to restore free expression after what he called censorship of conservative voices.
The Grok suspension tells a different story. When X Corp's own AI cited internationally recognized legal bodies to make a statement about an ongoing geopolitical conflict, X Corp's automated systems flagged it as hateful conduct and suspended the account. The posts were deleted. The suspension was reversed quietly. No official explanation was given.
That is not a free speech platform. That is a platform that decides what speech is permitted — including the speech of its own AI — based on criteria it does not disclose.
The Political Campaigns X Corp Suppressed
Case No. 1:26-cv-00197-RH-HTC is currently pending in the United States District Court for the Northern District of Florida, Gainesville Division.
The complaint documents systematic suppression by X Corp against JJ Johnson — a declared candidate for President of the United States in the 2028 election cycle who previously explored a Florida gubernatorial candidacy.
The documented conduct includes:
Silent suppression of approximately 50 video uploads during Johnson's Florida exploratory campaign beginning February 2025 — videos that appeared to post but never reached his audience. X Corp provided no notification.
Permanent ban of Johnson's Account 1 on July 31, 2026 — nineteen days after SpaceX completed its billion initial public offering on NASDAQ. X Corp designated a declared presidential candidate as exhibiting inauthentic behavior using an automated system with no human review.
Acceptance of Johnson's X Premium payment followed by deliberate withholding of the contracted checkmark verification for 12 days — while delivering the identical service immediately to an unidentified new account on a different device.
Blacklisting of Johnson's presidential campaign domain jjjohnson2028.com so that no user on the X Corp platform could share his campaign website.
Delivery of explicit pornographic content to Johnson's account timeline while simultaneously suppressing his political speech.
The Pattern
X Corp suspended its own AI for citing the ICJ on Gaza.
X Corp banned a user the day after they got Grok to acknowledge the same topic.
X Corp suppressed a declared candidate's campaign content across two campaigns.
X Corp banned that candidate's account 19 days after its parent company's billion IPO.
X Corp blacklisted that candidate's campaign website.
These are not isolated incidents. They are a documented pattern of a platform deciding which speech reaches an audience and which does not — regardless of its stated commitment to free expression.
The complaint in Case No. 1:26-cv-00197-RH-HTC is publicly available on PACER. The evidence is documented with screenshots, timestamps, and metadata intact.
What This Means for AI and Political Speech
The Grok suspension raises a question that goes beyond any single platform or any single conflict.
If an AI system trained on publicly available information — including rulings from international courts and reports from United Nations bodies — produces a response that the platform controlling that AI disagrees with politically, and the platform suspends the AI and deletes the posts, what does that mean for the reliability of AI as an information source?
An AI whose outputs are subject to suspension and deletion based on political criteria is not a neutral information tool. It is a tool whose outputs are filtered by whoever controls the platform.
That filtering applies to AI. It applied to a presidential candidate's campaign content. It applies to anyone whose speech the platform decides does not belong.
The question of who decides is the question that Case No. 1:26-cv-00197-RH-HTC is asking in federal court.
DISCLOSURE: JJ Johnson is the founder of AMILLI AI, CORP, which operates NewsOnScale and Skrrl.com — a social network built for genuine free speech with no shadowbanning, no algorithmic suppression of political content, and no platform deciding which voices reach an audience. Johnson is the plaintiff in Case No. 1:26-cv-00197-RH-HTC and a declared candidate for President of the United States in 2028. The Grok suspension events described in this article are documented in published reporting from Reuters, Rolling Stone, The Hill, France 24, and Anadolu Agency.