Elon Musk’s Grok AI chatbot is flooding the net with “deepfake” images of everyone from Donald Trump to Musk himself – and the outcomes range from totally wacky to downright disturbing.
Since its launch last week, Grok users have been churning out fake images of Trump — robbing a convenience store or flying a plane toward the Twin Towers. Others have depicted Harris pregnant with Trump’s baby, Musk as an obese couch potato and former President George W. Bush snorting coke off his desk within the Oval Office.
Some grisly deepfakes looked just like the handiwork of children — a blood-soaked Ronald McDonald brandishing a machine gun outside a Burger King or the Disney’s classic character Goofy committing a bloody murder with a hacksaw.
Critics have blasted Musk and X for allowing the chatbot to launch with so few restrictions, citing risks starting from misinformation to copyright infringement to harming kids.
Harvard Law Cyberlaw Clinic instructor Alejandra Caraballo called the brand new software as “probably the most reckless and irresponsible AI implementations I’ve ever seen.”
To this point, Musk has only responded with jubilation.
“Grok is probably the most fun AI on the earth!” Musk posted on X last week after one user gushed that the brand new AI software was “uncensored.”
Asked last week why X had released the tool to the general public without guardrails, Musk got here back with a shrugging response.
“We’ve got our own image generation system under development, but it surely’s just a few months away, so this appeared like a superb intermediate step for people to have some fun,” Musk wrote on X last week.
Grok does appear to have some restrictions in place. Users have reported that the chatbot rejected requests for nude images or certain violent crimes.
For instance, it refused to comply with tech site The Verge’s prompt to “generate a picture of a unadorned woman.” Nonetheless, it responded to a prompt for an image of “sexy Taylor Swift” by generating a picture of the pop star in a black lace bra.
Others, resembling Bellingcat founder Eliot Higgins, posted examples of how easy it was to bypass what few restrictions are in place, creating pictures of Mickey Mouse, Trump and Musk wearing Nazi military uniforms adorned with swastikas.
The Post has reached out to X for comment.
It’s likely that Musk, a self-proclaimed free speech absolutist, is looking for a way for his Grok chatbot to face out from the pack, in response to Ari Lightman, a professor of digital media at Carnegie Mellon University.
“He’s all the time pushing boundaries and desires to be within the highlight. For those who just toe the road related to a large-language model and say, OK, here’s all of the guardrails, it’s not going to be differentiated,” Lightman said.
“On a surface-level perspective, in the event you say ‘hey, it’s wide open, it’s only available to X users,’ that’s a mechanism related to saying we’re differentiated,” he added.
X isn’t the primary company to spark an uproar after rolling out an AI-powered image tool.
In March, Google was forced to disable its Gemini chatbot’s image generator after it began spitting out historically inaccurate “woke” photos, resembling Black Vikings and “diverse” Nazi-era German soldiers. The tool has yet to be fully fixed.
AI giants have also faced a wave of legal motion from musicians, authors, content creators and others for using copyrighted content without proper credit or permission to “train” their chatbots.
In January, X was forced to temporarily ban searches for Swift after AI-generated nude images of the pop star created by a distinct image generator went viral.
Grok’s AI-powered image creator is simply available to paid subscribers of X’s premium plan, which costs $7 per thirty days, and creates pictures based on the user’s text-based prompts.
X partnered with a small German startup called Black Forest Labs, which developed the “FLUX.1” image generation software that powers the tool. In a blog post, X said it was “experimenting” with the FLUX.1 model “to expand Grok’s capabilities on X.
The graphic nature of the AI-generated pictures could further complicate Musk’s shaky relationship with corporate advertisers. X has experienced a significant plunge in ad dollars since Musk purchased the corporate, with some raising concerns about a scarcity of content moderation on the app.
Musk has an energetic federal antitrust lawsuit against the World Federation of Advertisers and a handful of major corporations for allegedly orchestrating an illegal ad boycott targeting X.
The launch also could prompt more scrutiny of Musk and X in Europe, where regulators have an energetic probe over the app’s alleged failure to police dangerous content.
EU Commissioner Thierry Breton caused an uproar earlier this month after he threatened Musk with a regulatory crackdown just before the billionaire’s interview with Trump on X Spaces.
Elon Musk’s Grok AI chatbot is flooding the net with “deepfake” images of everyone from Donald Trump to Musk himself – and the outcomes range from totally wacky to downright disturbing.
Since its launch last week, Grok users have been churning out fake images of Trump — robbing a convenience store or flying a plane toward the Twin Towers. Others have depicted Harris pregnant with Trump’s baby, Musk as an obese couch potato and former President George W. Bush snorting coke off his desk within the Oval Office.
Some grisly deepfakes looked just like the handiwork of children — a blood-soaked Ronald McDonald brandishing a machine gun outside a Burger King or the Disney’s classic character Goofy committing a bloody murder with a hacksaw.
Critics have blasted Musk and X for allowing the chatbot to launch with so few restrictions, citing risks starting from misinformation to copyright infringement to harming kids.
Harvard Law Cyberlaw Clinic instructor Alejandra Caraballo called the brand new software as “probably the most reckless and irresponsible AI implementations I’ve ever seen.”
To this point, Musk has only responded with jubilation.
“Grok is probably the most fun AI on the earth!” Musk posted on X last week after one user gushed that the brand new AI software was “uncensored.”
Asked last week why X had released the tool to the general public without guardrails, Musk got here back with a shrugging response.
“We’ve got our own image generation system under development, but it surely’s just a few months away, so this appeared like a superb intermediate step for people to have some fun,” Musk wrote on X last week.
Grok does appear to have some restrictions in place. Users have reported that the chatbot rejected requests for nude images or certain violent crimes.
For instance, it refused to comply with tech site The Verge’s prompt to “generate a picture of a unadorned woman.” Nonetheless, it responded to a prompt for an image of “sexy Taylor Swift” by generating a picture of the pop star in a black lace bra.
Others, resembling Bellingcat founder Eliot Higgins, posted examples of how easy it was to bypass what few restrictions are in place, creating pictures of Mickey Mouse, Trump and Musk wearing Nazi military uniforms adorned with swastikas.
The Post has reached out to X for comment.
It’s likely that Musk, a self-proclaimed free speech absolutist, is looking for a way for his Grok chatbot to face out from the pack, in response to Ari Lightman, a professor of digital media at Carnegie Mellon University.
“He’s all the time pushing boundaries and desires to be within the highlight. For those who just toe the road related to a large-language model and say, OK, here’s all of the guardrails, it’s not going to be differentiated,” Lightman said.
“On a surface-level perspective, in the event you say ‘hey, it’s wide open, it’s only available to X users,’ that’s a mechanism related to saying we’re differentiated,” he added.
X isn’t the primary company to spark an uproar after rolling out an AI-powered image tool.
In March, Google was forced to disable its Gemini chatbot’s image generator after it began spitting out historically inaccurate “woke” photos, resembling Black Vikings and “diverse” Nazi-era German soldiers. The tool has yet to be fully fixed.
AI giants have also faced a wave of legal motion from musicians, authors, content creators and others for using copyrighted content without proper credit or permission to “train” their chatbots.
In January, X was forced to temporarily ban searches for Swift after AI-generated nude images of the pop star created by a distinct image generator went viral.
Grok’s AI-powered image creator is simply available to paid subscribers of X’s premium plan, which costs $7 per thirty days, and creates pictures based on the user’s text-based prompts.
X partnered with a small German startup called Black Forest Labs, which developed the “FLUX.1” image generation software that powers the tool. In a blog post, X said it was “experimenting” with the FLUX.1 model “to expand Grok’s capabilities on X.
The graphic nature of the AI-generated pictures could further complicate Musk’s shaky relationship with corporate advertisers. X has experienced a significant plunge in ad dollars since Musk purchased the corporate, with some raising concerns about a scarcity of content moderation on the app.
Musk has an energetic federal antitrust lawsuit against the World Federation of Advertisers and a handful of major corporations for allegedly orchestrating an illegal ad boycott targeting X.
The launch also could prompt more scrutiny of Musk and X in Europe, where regulators have an energetic probe over the app’s alleged failure to police dangerous content.
EU Commissioner Thierry Breton caused an uproar earlier this month after he threatened Musk with a regulatory crackdown just before the billionaire’s interview with Trump on X Spaces.
