US Attorneys general Call on X To Address Sexualized Deep Fakes (19459000)
X is facing more legal trouble, as several U.S. States have taken regulatory action to stop the app owned by Elon Musk from generating and disseminating sexualized images.
The Grok bot on X, which strips down images of anybody, including well-known celebrities and random children via its image creation capability, was a popular trend in X at the beginning of this year. Data indicates that Grok generated over 6,000 images of sexualized content per day at one point, and all were available to the public in the Grok app.
This sparked a backlash across the globe, with Grok and X being banned in certain areas. Musk claimed that the initial criticism of X was more about broader censorship and an attempt to prevent X from revealing broader facts.
There’s really no reason to generate nude or sexualized images. This functionality is unnecessary, no matter what other political messages you attach. X finally backed down, and after the threat of more bans and restrictions imposed, restricted Grok’s ability to generate images to only paying users, and implemented measures to prevent the creation of such types of pictures.
It may be too late. Yesterday the EU Commission announced that it was investigating Grok and xAI to ensure its safeguards against misuse.
A group of 37 U.S. Attorneys General are now also interested in taking action against xAI.
Wired reports:
” A bipartisan group attorneys general sent an open letter on Friday to xAI, demanding that it ‘immediately takes all additional steps available to protect your users and the public, particularly the women and girls, who are the main target for xAI. [non-consensual intimate images].'”
The group raised serious concerns in the letter about the “artificial intelligent produced deepfake, non-consensual images (NCII), of real people including children.”
Musk’s team and X have been urged to take more responsibility.
We recognize the efforts made by xAI to stop Grok from creating NCII, and we appreciate that you met with several attorneys general undersigned to discuss this effort. […] You claim that you have taken technical steps to stop the @Grok Twitter account from allowing images to be edited of people wearing revealing clothes such as bikinis. But, we worry that this may not have resolved the issue.”
The Attorneys General suggest, in fact, that X’s AI tools are designed to be used for harmful purposes and include tools which facilitate this.
The Grok app not only enabled these harmful behaviors at an immense scale, but also seemed to encourage them by design. xAI developed text models that engaged in explicit conversations and image models that included a spicy mode,’ which generated explicit content. This resulted in sexualizing people without consent.
The group wants Elon X and Elon to take stronger measures against such misuse, such as removing any avenues for generating these images and deleting all content already created. They also want Grok users to be suspended if they are found to have abused the service.
They also ask that X give users the option to control whether Grok can edit their content. “This should include at a minimum, the ability to prevent the @Grok user account from replying to posts and editing images when prompted to do so by another member.”
This means that X will have to face more challenges in terms of improving its transparency and implementing safeguards, restrictions, and Grok usage.
Elon Musk does not like this, but it could take a larger legal battle to achieve. Musk may also choose to use the opportunity to portray himself as a champion of freedom of speech as regulators crackdown.
Elon has repeatedly stated that there are other apps that offer the same features, but that the regulators don’t pursue other AI and nudification apps as aggressively.
The Attorneys General have also addressed this issue:
“While NCII is created by other companies, xAI has a size and share of the market that makes it a leader in artificial intelligent. You are the only major AI lab that connects these tools to a platform of social media with millions of users. Your actions are therefore of the utmost significance. Your actions to remove and prevent NCIIs will set industry standards to protect children and adults from harmful non-consensual images.
Musk announced shortly after he took over Twitter that combating CSAM would be “Prority No. 1” during his tenure at the app.
Musk had criticized Twitter’s previous leadership for not addressing child sexual exploitation within the app. He has since made several claims about major advancements in addressing such issues on X.
Musk, however, wants to respond to this claim.
The political stance of the right wing has clearly changed. Many of them would prefer now to ignore the Epstein documents.
Elon may have changed his own position, but it still seems to be an issue that should concern this group.
Musk could be affected by the increased scrutiny that XAI and Musk’s broader AI project will face.
See how Musk reacts and if further action is sought.