No, Grok hasn’t paywalled its deepfake picture function

Elon Musk’s X has partially restricted entry to Grok’s picture modifying capabilities amid rising backlash to the flood of nonconsensual, sexualized deepfakes of adults and minors generated by the platform. As of this writing, it now not generates photographs as @grok replies at no cost, however Grok’s picture modifying instruments stay available for any X consumer to churn out photographs, each sexualized and tame.

X customers had been beforehand in a position to ask Grok – by tagging @grok in a tweet – to edit or create photographs on the platform by tagging it in a submit. Customers now making an attempt this are met with an automatic response from the chatbot’s account, telling them that “Picture technology and modifying are at the moment restricted to paying subscribers.” The reply features a hyperlink encouraging them to subscribe to X’s paid applications “to unlock these options.”

The message gives the look that solely paid X customers can edit or create photographs utilizing Grok. That message is bolstered within the many headlines circulating proper now stating that xAI, one other Musk firm that makes Grok, has restricted Grok’s picture modifying talents to this choose group of customers. However that impression is fake. All X customers — together with free ones — can nonetheless use Grok to edit and create photographs, together with the sexually suggestive deepfakes that landed Grok in sizzling water to start with.

Interacting with Grok by replies on X is only one of a number of methods to make use of the AI chatbot. For instance, the “Edit picture” button nonetheless invokes Grok on each picture displayed on X’s desktop web site, and that very same button will be accessed on the X app by long-pressing on any picture. Grok may also be accessed by a standalone web site or app, in addition to through a extremely promoted tab within the X apps and X web site. These stay available to free customers, and require no extra effort in comparison with the @grok command.

The Verge examined the Grok web site, app, and X tab with free accounts and Grok readily complied with our requests to edit or generate a wide range of photographs, together with a full “nudify” and one in all Musk sporting a bikini.

The restrictions on X come after Grok has been used to generate sexual photographs primarily of ladies, a lot of them actual, in addition to minors. A lot of this landed on X, which the Monetary Occasions dubbed “the deepfake porn website previously often called Twitter.” The deepfake undressings, in addition to different sexual materials, have outraged regulators worldwide, who’ve threatened to take motion towards X, although the deepfakes could fall into one thing of a authorized void.

Musk and xAI have additionally threatened to take motion towards customers who create unlawful content material utilizing Grok, reasonably than guaranteeing that guardrails exist to make it unimaginable to create it within the first place.

X’s strategy — limiting entry reasonably than constraining use — contrasts with the strategy taken by different AI firms like Google and OpenAI, which are likely to impose strict guardrails on instruments like Gemini and ChatGPT, reasonably than permit free rein for individuals who pay. Musk has reportedly opposed stricter guardrails personally, and a number of other members of xAI’s already sparse security workforce reportedly stop within the lead as much as the deluge of digital undressings.

X didn’t reply to The Verge’s request for remark.