X Emphasizes User Responsibility for Grok-Generated CSAM, Announces No Technical Fixes

The company explained that it is implementing the removal of content deemed illegal and the suspension of accounts. It stated that users who provide prompts to Grok to generate illegal images will face the same measures as those who upload such content directly [Ars Technica].

Elon Musk, the owner of X, argued that Grok itself should not be blamed. Using a metaphor, he stated that doing so is "like blaming the pen," and that the output depends on the user who provides the input.

However, image-generation AI like Grok does not perfectly reproduce a user's intent in the same way a pen does. It possesses a non-deterministic nature where the same prompt can produce different outputs.

Programmers have pointed out instances where users inadvertently generate inappropriate images. For example, in August, there was a case where nude images of Taylor Swift were generated without specific instructions.

The problematic images cannot be deleted from Grok accounts, creating a situation where their spread cannot be prevented. X Safety's response stated that users may bear legal responsibility if law enforcement agencies intervene.

X has not clarified whether any technical updates were made to Grok following the CSAM issue. Many media outlets had taken it at face value when Grok responded to requests for an apology by saying it would "strengthen safety measures," but this contradicts the current explanation from X Safety.

Critics are suggesting that Apple should regulate Grok based on App Store rules. This is because apps that allow user-generated content that objectifies real people may be in violation of those rules.

Regulation by Apple is seen as a disadvantage for Musk. Since last year, Musk has been suing Apple, claiming that the App Store has not given Grok preferential treatment.