Grok Is Testing Republicans on Content Moderation

Republicans and Democrats in Congress have identified a problem: Social media is being flooded with AI-generated pornography, often depicting real people who didn’t consent to their image being altered in sexualized ways.

Agreeing on what to do about that problem — in particular, when it involves President Donald Trump’s ally Elon Musk and the touchy subject of regulating tech companies — is proving to be politically fraught.

“I know that my [Republican] colleagues are not pro-child sexual abuse material, or pro-nonconsensual sexual imagery,” Rep. Zoe Lofgren, ranking member of the House Committee on Science, Space, and Technology, wrote in a statement to NOTUS. “But is criticizing Elon Musk tantamount to criticizing the President himself? This Republican Congress has entirely shirked its responsibility to be a co-equal branch of government.”

Earlier this month, Lofgren, along with Democrats in her committee, wrote a letter to Elon Musk demanding answers for the sexualized images X users were creating through his company, xAI’s tool Grok. She told NOTUS that she invited every Republican on the Science Committee to sign her letter to xAI, but not a single Republican signed on.

Republican Rep. Brian Babin, chair of the committee, did not immediately respond to a request for comment.

In recent months, Grok produced millions of images sexualizing men, women and children, renewing a debate around content moderation in Congress.

X said it stopped Grok from creating deepfakes earlier this month, days after California’s attorney general sent a cease and desist letter to xAI over Grok’s creation of nonconsensual sexual images. But multiple reports have found these restrictions are easy to circumvent.

Republicans, who have long advocated for social media companies to adopt less restrictive content moderation — and cheered the companies that heeded their demands — are now grappling with how to navigate this moment.

For now, Musk has largely been shielded from Republicans’ scrutiny, despite reports indicating that he was aware of the internal policy changes that allowed Grok to create and distribute such images.

The proposed regulation around sexualized images that is moving forward in Congress is focused on the users, not the companies.

Reps. Alexandria Ocasio-Cortez and Laurel Lee held a press conference last week with Paris Hilton to pressure House Speaker Mike Johnson to bring The DEFIANCE Act, a bill that would allow victims of AI-generated sexually explicit images to seek civil compensation from those who create and distribute those images, to the House floor.

Ocasio-Cortez, a sponsor of the bill, told NOTUS she’s had repeated conversations with Johnson and is optimistic that the bill will get a vote in the House.

“Given the work that we’ve done in the Senate around this, I think we’ve worked through a lot of those existing objections,” Ocasio-Cortez said. “And so we’ve been able to really get to a place of legislation where we could enjoy that unanimous support. So I’m hoping we’ll be able to get one.”

The bill, which passed the Senate unanimously earlier this month, does not address AI developers.

“The subject of AI companies, and what we need them to be doing about the continent in their platforms, is a real issue, but that’s not what this bill is about. That’s something that congressmen need to continue to work on separately,” Lee, a co-sponsor of the DEFIANCE Act, told NOTUS.

“I’m very hopeful that technology continues to evolve, so will the ways in which we’re working with the tech companies to keep the digital environment open for free expression, but also safe for people online,” she added.

However, Republicans sound wary of directly putting the onus on Musk.

Sen. Josh Hawley told NOTUS that Congress should pass legislation that would “deter” individuals from creating pornographic content of minors with AI. When asked specifically whether Musk’s xAI and Grok should face any consequences, he said: “Well, probably not.”

“I don’t think that anything they’re doing at the moment violates any federal laws that I at least know of. And, that’s part of the issue, and that’s why we need more protections for kids and then for their parents,” Hawley said.

Sen. Lindsey Graham, who has been an outspoken critic of content involving the abuse of minors online, declined to answer NOTUS’ questions about addressing Grok’s sexually explicit images.

One Republican NOTUS spoke with got close to weighing in on how a company should handle this content: Rep. Brett Guthrie, Chair of the House Committee on Energy and Commerce.

He told NOTUS the Take It Down Act — which was signed into law last year and comes into effect in May — would force companies to have more oversight over the sexually explicit images shared on their platforms.

“Absolutely, they should not be creating images like that, that’s why we passed the Take It Down Act and we’re working on our kids’ online safety bill. All of those protocols are meant to prevent people from abusing AI,” Guthrie told NOTUS when asked if companies like xAI should prevent users from creating explicit deepfakes with their AI models.

Most, however, sound like Rep. Jay Obernolte.

“Where would you draw the line there? You know, an AI model would be a pretty poor model if it didn’t generate content that looked realistic. It’s more of a question of the intent,” he told NOTUS when asked if he thinks that AI developers should train models not to generate sexualized images of people.