Elon Musk's Xai has announced a new feature in the AI chatbot Grok that allows users to generate 6-second video clips containing those of a more distinctive nature.
This feature, called “Imagine,” creates a video with sound according to user prompts, according to an X post on the official Grok account. A thread from Xai employee Mati Roy (Shard by Musk) highlighted examples of videos produced by the Grok update, including robots, “Alien Tribe Women.” The posted thread roy was labeled as adult content in X.
According to NBC News, Roy also said, “Grok Imagine Videos has a spicy mode that allows you to do nude.” (The original tweet appears to have been deleted, but other tweets have also flagged the existence of “spicy mode.”
Imagine's release date is set to October. While some employees and subscribers seem to have early access to the feature, Musk shows that programmers are still thinking about final adjustments.
The latest update is part of Grok 4, the latest version of the AI chatbot that the company features as “the world's smartest artificial intelligence.”
However, in recent weeks, launches related to GROK have been smeared with controversy, including the introduction of sexual AI “companions.” And this new feature raises unique concerns about its potential use in the creation of deepfakes. This changed the video of an individual that can mimic someone's portrait and is often published without consent.
Instead of listening to the call to remove the “NSFW” AI chatbot, Xai appears to be doubling its way by encouraging AI videos to create nudes.” “It has not been confirmed that they will not create pornographic content similar to recognizable people. Xai should look for ways to prevent sexual abuse and exploitation.”
Xai did not immediately respond to Time's request for comment.
Here's what you need to know about new developments:
Deepfake concerns
The latest Xai update arrives amid public concerns about deepfakes. According to a 2019 poll from the Pew Research Center, roughly three-quarters of US adults are in favor of limiting the use of digitally modified videos and images.
A recent January poll from the Institute for Artificial Intelligence Policy found that 84% of Americans support laws that make uncreative deepfark porn illegal.
Roy's recent post on Grok's AI video generator said the service can create realistic videos of humans and animate static images. Some people are noting that new services could be a way to generate personalized porn images.
Several people have moved forward with concern about Grok being used to generate sexually explicit images within portraits, even before the recent update. The young woman Evie was identified on USA Today said that using an AI chatbot, her selfie was transformed into an explicit sexual image by others. “It was just a shock to see a bot built into a platform like X being able to do that,” she said.
X has moved to block the image from using certain phrases and words, according to the outlet, but online users have found other ways to provoke victims with chatbots, such as asking them to generate detailed stories of Evie's sexual assault and rape.
In May, the president signed The Take It Down Act. This requires a platform to “quickly remove” such unconsensual intimate visuals.
The law aims to protect minors and adults who may have digitally changed their image during a situation of “reasonable expectations of privacy.”
Before the law passed, Democrats moved to restrict images created by Grok, particularly in August 2024, sending a letter to the Federal Election Commission, calling for restrictions on the use of deepfakes in election candidates. The letter pointed to the deepfakes of 2024 Democratic presidential candidate Vice President Kamala Harris and pop star Taylor Swift.
Other recent Glock controversies
In such an example, due to the update, the chatbot issued an anti-Semitic statement, calling itself a “Mechahitler,” urging the company to apologise and take steps to address the issue. Grok was temporarily closed due to offensive statements made by the chatbot.
Another update earlier this month announced New Xai “Companions,” a set of characters with their own personas that Grok users can chat with. Users have expressed concern about the development of a sexually animated character named “Ani”. “Ani” speaks in a sultry tone, revealing the lingerie set following a set of flirty interactions, even when the app is in “kids mode.”
The character introductions have been criticized by some users and organizations, including the National Center on Sexual Exploitation, which pro-porn and sexual exploitation.
Mask said after announcing that “customizable companions” would also be available to users soon.
