|
Hey there, I am running out of ways to explain how bad Facebook has become. However bad your mental conception of Facebook is after reading this and other articles I have written about its AI spam problem, I assure you it is much worse than that. I can only convey a small amount of what's happening there, and 99 percent of the rot that I see on a daily basis does not make it into these stories because there's only so many hours in the day. Facebook's response to all of this is that it does not care and that it actually likes this sort of content because it makes the company money, so, good luck to us all. -Jason An AI-generated Elon Musk is standing in front of a tiny house on wheels, holding a key that says “TSSLA” that has a price tag of $6,999. The Facebook post, by a page called “Easy Recipes,” says “Elon Musk’s $6,789 Tiny House FINALLY HIT The Market!” “If successful, this affordable housing option could be more than just a way to reduce costs—it could represent a bold new approach to one of America’s most pressing challenges: affordable housing,” the post reads. “As property prices rise and wages stagnate, the dream of homeownership is slipping away for many. Musk’s plan for these accessible homes proposes a fundamental shift, making homeownership viable for first-time buyers and those facing financial hurdles.”  The post has nearly 2,000 comments, 500 shares, and 5,800 likes. Comments include “Elon Musk you're a life saver for millions of people.” “This man is making it possible for us not to have homeless people.” “He’s a great person.” “This would be good for seniors as well!” “This is a great plan! No more homeless people in America!” “God Bless him for sharing his wealth. Hope this helps the homeless and veterans and hurricane victims.” “Thank you Elon this could help a lot of people.” “Make America Great again and affordable.” “It is way better than tent cities for crying out loud.” “Elon is amazing! So is Trump!”
This segment is a paid ad. If you’re interested in advertising, let's talk.
Generative AI (GenAI) chatbots are changing the way businesses engage with customers by providing intelligent, round-the-clock support. However, this technology comes with serious risks, including the potential for unsafe responses and the mishandling of sensitive requests, which can jeopardize a company’s reputation and customer safety.
In ActiveFence’s latest report, we explore these vulnerabilities via a case study from the travel industry to highlight broader implications for business using GenAI-powered applications. It also examines the challenges faced by travel companies, including risks to brand integrity and safety concerns related to user recommendations. Plus, we provide practical strategies for improving safety measures and protecting your GenAI applications from potential threats.
You’ll also find real-life examples of how our researchers manipulated the AI systems with risky prompts to reveal unsafe advice.
Stay informed and safeguard your business—read the full report today.
Some of the most viral posts on Facebook leading up to and immediately after an election in which Musk actively campaigned for Donald Trump are AI-generated disinformation that universally make Musk look like a philanthropist and genius inventor who is actively solving America’s problems out of the kindness of his heart, created by people in the Philippines, Vietnam, and India and spammed to Facebook. These images went viral and continue to go viral on a platform that has repeatedly said it does not want anything to do with politics, gave less reach to posts that contained the word “vote,” and suspended journalists for reporting about Trump’s praise for Hitler’s generals. Dozens of these Elon Musk images were delivered directly to me on my news feed on a Facebook account I use to research AI spam. I found thousands of AI-generated Elon Musk images that are specifically masquerading as news being posted incessantly by dozens of pages. Like other Facebook AI spam I’ve reported on, many of these pages link offsite to ad-laden AI spam farms, but, more likely they are trying to make money directly from Facebook using the creator bonus program, which pays a fraction of the ad revenue generated from engagement on the posts. I have written extensively about AI-spam on Facebook and have researched whether the comments on this type of content are real, and with AI-generated Elon Musk spam in particular, the vast majority of comments appear to come from real people who are actively engaging with the substance of the posted content. PolitiFact, which is an official factchecking partner of Facebook, previously wrote about fake images of Musk promoting affordable housing on Facebook. Still, I found no factchecking or AI labels on any of the Musk spam I found. Some of the pages focus entirely on posting AI-generated inspiration porn about Musk, while others mix in Musk content alongside other celebrity news. One page, “American Spotlight,” has an avatar that just reads “AMERICA NEWS” and claims to be operating from Las Vegas. Page transparency information, meanwhile, shows the page’s administrator is in Vietnam. The posts about Elon Musk are a mix of things that have absolutely no basis in reality alongside warped versions of news articles that are kind of true, for example, there are some posts about SpaceX’s Starship rocket, but with AI-generated images and that tout features that do not exist.  One post by American Spotlight reads “Breaking news: Elon Musk Unveils UFO Fighter Jet That Defies Physics” with an AI Musk standing in front of an AI “TESLLA”-branded flying saucer. It has 1,100 likes and 439 comments. Comments include “This kinda stuff happens only in America,” “Right out of Starwars!!! LOL Awesome Technology!!! and “Amazed with your God’s gift for spaceX ability … I pray you make the right choices for president Trump. And thank you for leading trump rallies <3”  Another post from American Spotlight from before the election reads “Tesla CEO Elon Musk Releases First Flying Car That Changes Everything!” It has 4,600 likes, 1,000 comments, and 276 shares. A post from an account called StellarX posits that Musk has developed “Iron Man-Style Armor” as a response to “Life-Threatening Attacks” and “two assassination attempts” and feature various AI-generated images of him in Iron Man armor or standing among robots. It has 5,100 likes and 1,400 comments. Many of these comments are praying for his safety and say things like “thank you for your contribution to our world and for supporting trump” and “I pray that he will stay safe. He’s so very brilliant and I am personally looking forward to his joining President Trump to implement joint development in our future” and “Why do we never hear of assassination attempts on George Soros??? He has financed the worst evil this country has seen over the past two decades.” A page called Earth Guardian is posting Elon Musk AI-generated inspiration porn every several minutes, which in the last day has ranged from a “UFO fighter jet that defies physics” to a “supersonic space jet” to a “$15,000 motor home,” to the “FASTESTS Hypersonic Jet Reaching the speed of light,” and an “insane warp drive starship.”  I have reported on AI-generated spam for a year now, and have watched as different trends come and go. In August, when I wrote about this community of people, bizarre Jesus content, surreal landscapes and dream homes, and birthday celebration posts were performing very well on Facebook. At the time, one Facebook AI spammer told me that they intended to begin spamming Facebook with “American news.” It is clear that pro-Elon Musk inspiration porn is the new strategy, or “meta” for these spammers, and creating AI spam that specifically targets people in the United States is part of the new strategy. The YouTube page for one of the Indian influencers who teaches people how to do this is full of videos for “USA CHANNELS” and US-focused spam. This strategy is clearly trickling down to Facebook at a large scale that is impossible to quantify or systematically study because Meta has killed CrowdTangle, a research tool that showed how content spread on the platform.  Something new about the Musk spam on Facebook is that there are lots of AI-generated Reels and longform videos to go alongside the photo spam. For example, here’s an AI video with 240,000 views and 12,000 likes that says Musk will put “one million humans on Mars in the next two decades.” One page that is posting dozens of AI-generated videos about Musk called “Elon Musk Idea” was until recently posting AI-generated images meant to capitalize on suffering in Gaza and was recently called “have a nice day” and “cute wonderful lovely flower decoration,” which are names that have been commonly used by AI spam pages. The videos page features AI-generated videos that say Musk has predicted the end of America, that he has said “Russia will destroy us,” and, also, older selfie videos of a man who seems to run the page speaking Khmer, the official language in Cambodia. To see how many of these pages there are, look at the search results for “water engine,” which is one of the many fake AI-generated miracles that Elon Musk has supposedly created in the last several weeks: While this type of Elon Musk AI spam appears to be most prominent on Facebook at the moment, many of the people making it are clearly posting it to other platforms, as well. McKenzie Sadeghi, who studies AI and foreign influence at NewsGuard, found similar AI spam about Musk on YouTube, Threads, and TikTok when we showed her some examples. “It appears there was a spike in public interest for Elon Musk around the time of the U.S. presidential election, coinciding with the timing of some of these posts,” Sadeghi told 404 Media. “I'm not aware of the full extent/time frame of this campaign but it is possible that these accounts were attempting to capitalize on this surge in public interest by pumping out clickbait-like, AI-generated content of Musk, and in some cases, directing users to find an article about Musk in the pinned comment in an effort to bypass Facebook’s algorithmic limitations on external links.” Meta acknowledged a request for comment but did not say whether this type of content was allowed. Mark Zuckerberg recently said that AI generated content has increased Facebook and Instagram’s time-on-site and that more AI-generated spam will be encouraged. Meta has done absolutely nothing remotely meaningful to prevent this and continues to financially subsidize it.
|
