Home News AI Chatbots Are Making LA Protest Disinformation Worse

AI Chatbots Are Making LA Protest Disinformation Worse

0
AI Chatbots Are Making LA Protest Disinformation Worse

AI Chatbots Are Making LA Protest Misinformation Worse (19459000)

The spread of misinformation about the Los Angeles demonstrations is exacerbated by users using AI chatbots such as Grok and ChatGPT for fact-checking.

As LA residents took to the streets to protest the increasingly frequent Immigration and Customs Enforcement raids, conservative posters flooded their social media feeds with inaccurate news. In addition to the well-worn tactics such as Repurposing old footage of protests Clips from video games Moviesand posters have claimed that protesters are nothing more than paid agitators. There is no evidence that the alleged shadowy forces are behind this.

As companies like X, Meta, and others have stepped away from moderating content on their platforms and in the midst fast-moving, divisive stories like the LA demonstrations, users have turned to AI chatbots–which have been inaccurate in many cases.

The San Francisco Chronicle published a story on Monday. Published imagesshowing National Guard troops sleeping in floors. Later, California Governor Gavin Newsom shared them on X in response to a tweet from President Donald Trump. Writing – “You sent your soldiers here without fuel, water, food or a place for them to sleep.”

Within a few minutes of the posts’ being shared, users on X or Facebook claimed that the images had either been AI-generated or taken in a completely different scenario.

Laura Loomer, a conspiracist on X, claimed that @GavinNewsom had used an AI photo to smear Trump. Grok, X’s chatbot, was contacted by some users in search of clarity. One user asked it to clarify the location of the photo.

Grok wrote, “The photos are likely to have originated in Afghanistan in 2021 during the National Guard evacuation efforts for Operation Allies Refuge.” Claims linking them with the 2025 Los Angeles deployment are not credible and appear to be an incorrect attribution. Grok responded to a question from another X user, who said that the San Francisco Chronicle had first obtained the images. The photos of National Guard soldiers sleeping on floors likely date from 2021. They are probably taken at the U.S. Capitol and not Los Angeles in 2025.

The San Francisco Chronicle didn’t immediately respond to an inquiry for comment.

Grok may not be the only tool that could make the situation worse. Melissa O’Connor describes herself as “OSINT Citizen Journalism.” ChatGPT gave herresults when she uploaded Newsom’s pictures of soldiers sleeping on the ground. OpenAI’s Chatbot incorrectly stated

that one of the photos shared by the California Governor was taken at Kabul Airport in 2021, during former President Joe Biden’s Afghanistan withdrawal. The same results have been shared on other platforms, such as Facebook and Trump’s Truth Social, as proof that the photos are fake. (O’Connor later (She posted in order to let her know that she now knows the photos are not older than four years, even though the original post was still left up.

LeadStories fact-checked the image and found that it was taken in a New Jersey suburb. Grok, when asked to clarify where the image came from, wrote: “The photo is likely a real photo from Paramount, Los Angeles, taken on June 7, 2025, near the Home Depot on Alondra Boulevard during protests against ICE raids.” His post has now been viewed nearly 4 million times.

LeadStories fact-checked the image and found that it was taken in a New Jersey suburban area. When Grok was asked for clarification, it responded: “The photo is likely a true photo taken in Paramount, Los Angeles on June 7,2025, near a Home Depot on Alondra Boulevard, during protests against ICE Raids.”

Another X user pointed to the fact that the image had been taken in New Jersey, and asked Gro ABC7, Los Angeles Times and other news reports confirm that bricks were thrown during clashes between federal agents.

WIRED was unable to identify any reports from the mentioned outlets indicating bricks were thrown in the recent protests.

X, the operator of ChatGPT and OpenAI did not respond to requests for comments immediately.

Chatbots’ unreliability is adding to a disinformation landscape that has become so common on social media in the wake of major breaking news.

Ted Cruz, a US senator from Texas, quoted a Woods post on Sunday night. He wrote: “This… is… not… peaceful.” Woods shared a video that has been deleted by its original poster. It was taken at the Black Lives Matter demonstrations in 2020. Cruz and Woods, despite the millions of views, have not taken down their posts.

On a Monday evening, a tired conspiracy theory popular among right-wingers surfaced. Many pro-Trump accounts claimed that protesters had been paid shills, and that shadowy but largely unspecified individuals were funding the entire thing. This narrative was sparked when news footage showed people handing out face masks called “bionic shields” from the back of black trucks.

Right-wing YouTuber Benny Johnson posted on X that “Bionic face masks are now being distributed in large numbers to rioters in Los Angeles.” He added “Paid insurrection.”

www.aiobserver.co

NO COMMENTS

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Exit mobile version