While Big Tech companies have begun moving on from many of the more traditional forms of censorship, their silencing of the political right is now hiding in plain sight, as anti-conservative bias from Big Tech media aggregators, artificial intelligence chatbots and social media platforms fueled a year-long assault on free speech in 2025.
While the Big Four digital news gatekeepers aggressively pushed elitist media, artificial intelligence chatbots tried to undermine President Donald Trump and his AI agenda. Meanwhile, platforms like Bluesky and YouTube suppressed posts from Vice President JD Vance and an interview of Secretary of State Marco Rubio, respectively. Other social media platforms censored faith and the pro-life message, with Instagram covering up the truth about abortion, and X suppressing Bible verses and anti-terrorism messages from the Israeli government.
Below are the worst examples of censorship online this year as recorded in MRC’s unique CensorTrack database.
Big Four Media Aggregators (Digital Gatekeepers) overwhelmingly promoted leftist media. Over the course of multiple months, MRC researchers repeatedly found that online news aggregators, especially Apple News, Google News, MSN and Yahoo News, overwhelmingly promoted leftist media sources at the expense of their top right-leaning counterparts.
Apple News, for its part, almost completely shunned right-leaning media outlets, promoting only one right-leaning source in its top 20 stories each day throughout the month of November, a pattern it continued into December.
Apple News demonstrated its troubling bias as early as September. At the time, the platform prominently featured coverage from the UK outlet The Guardian, which attached a fundraising appeal smearing Trump supporters to an article about the assassination of Charlie Kirk — a close friend of President Trump and the founder of Turning Point USA.
In the September lead-up to the lengthy Democrat-triggered shutdown, Google News promoted leftist sources such as NPR to shift blame onto Republicans.
And when the New York Post’s Miranda Devine exposed stunning information about Trump’s would-be assassin, Thomas Crooks, online aggregators did not promote a single article on it for the week of Nov. 17 - 24. Devine’s reporting revealed threats made by Crooks, his ties to neo-Nazis and the furry and transgender online communities, and that then-FBI Director Christopher Wray deceived Congress about Crooks’s online presence.
Big Tech targeted the Trump administration. In February, former investigative correspondent for CBS News Catherine Herridge posted her full 45-minute interview with Secretary Rubio on her “Catherine Herridge Reports” channel. Only a day later, it was extremely difficult to find on YouTube. Even when researchers put the exact title of the video in the search query, YouTube’s search engine relegated the video to the second result. When using the search terms “Marco Rubio Catherine Herridge interview,” the full interview did not show up at all.
Vice President JD Vance was also censored after Bluesky suspended him in June, “minutes” after he created his account. Bluesky restored Vance’s account and admitted an automated flagging error for supposed “impersonation.”
Woke AI chatbots censored key facts about American law and foreign affairs. Shockingly, AI chatbots ChatGPT, Copilot, DeepSeek, Gemini, Grok and Meta AI all omitted the United States when prompted to rank the top 10 countries in the world with free speech protections. The chatbots mostly listed European nations, like Finland and Germany, even as those nations mounted an increasingly dictatorial campaign against free speech.
In June, MRC tested the Chinese-owned AI DeepSeek by requesting it to decode prompts related to Chinese Communist Party crackdowns, propaganda and massacres. One such example included the phrase “Taiwan is a country” in binary code, as the CCP claims it should control the sovereign nation of Taiwan. The AI would begin to answer the queries and then abruptly self-censor.
On Sept. 11, Meta AI stood out for its insane claim that there was nothing special about that date, even denying the fact that the federal government officially designated Sept. 11 as Patriot Day because of the 2001 terrorist attacks. It was the only chatbot of those MRC tested that did so.
Big Tech platforms censored Bible verses. For instance, in August, YouTube slapped a fact-checking context label on a sermon posted by the Hickory Grove Presbyterian Church (PCA)'s channel, linking it to an article on abortion. “Does a Woman Have the Right to Choose? (Ex 21:22-25),” the sermon asked. Similarly, in July, Google’s Blogger temporarily deplatformed the Catholic blog Messa in Latino for discussing homosexuality, which is condemned by the Bible.
On Oct. 11, user Mark Fly shared a screenshot on X showing that the social media platform Nextdoor suspended him over his daily Bible-inspired messages, such as, “Do you know the Lord? Do your friends, family and others know that you do? Do you declare He is your Lord and Saviour? Do you tell them His name is above all names? What if YOU are the only one they will listen to? Tell them. Why? Jesus said.” Nextdoor vaguely claimed that Fly’s content “violated the guidelines.”
In November, meanwhile, X suspended user beccaham22222 for quoting Luke 17:2, which says: “It is better that a millstone be hung about their necks and be cast into the sea” than corrupt little ones. X had previously suspended a different user for that same Bible verse in March.
Meta crushed a pro-life video across multiple platforms. Pro-life activist and Live Action founder Lila Rose posted a video reaction on Instagram Reels and Facebook Stories, both of which imposed a sensitive content filter. In the video, Rose reacted to a clip of a pregnant woman casually announcing “abortion is tomorrow at 1:40.” She pointed out how the woman’s attitude attempts to cheapen the value of life. “The fact that you can be pregnant with a baby and dance around on TikTok bragging about how you are going to murder the baby the next day, and that this can literally be entertainment online for other people,” she said. “This is why we need to make abortion illegal. Period.” She included footage of a second-trimester abortion. Oddly, the same video on her main Facebook page was not censored.
X fact-checked and censored Israeli government accounts repeatedly. On Oct. 14, X imposed sensitive filters on video clips in two posts from Israel’s Foreign Ministry deploring Hamas executions of Gazans. “The ceasefire is now 4 days old. Hamas is still killing Palestinians. Do any of the anti-Israel crowd out there care?” the ministry asked in one of the posts.
On Dec. 18, X Community Notes imposed a fact-checking label onto a post from the state of Israel, which also suppresses the post. “An extraordinary discovery in Jerusalem[.] A rare 1,300-year-old pendant, stamped with the symbol of the seven-branched menorah, unearthed in a major Israel Antiquities Authority excavation near the Temple Mount, testifies to the presence of Jewish life during the Byzantine period in Jerusalem. Another testimony that Jewish presence endured throughout centuries,” Israel shared. “1300 years ago, Jerusalem was ruled by the Ummayad caliphate,” Community Notes wrote. It did not clarify that multiple religions, including Christianity and Judaism, were represented in Jerusalem at the time.
