A recent update to Elon Musk's Xai Chatbot Grok has launched two new “companions” or AI characters for users to interact with. This includes a sexually blonde anime bot called “Ani” that users can access even if the app is in “kids mode.”
The new version of GROK allows users to interact with AI as if they were talking to certain characters. One of the characters known as “Bad Rudi” is a Red Panda, programmed to sham users in graphic or vulgar ways, but can turn off their personality traits. (“Companion” may also be referred to by Grok as “Bad Rudy.”)
The other is “Ani”. A young woman with a black corset, fishnet tights and a black dress that is short off her shoulders, tightly choked by Lacy's choker, responds to the prompts with a slow, sultry voice.
The character is powered by Grok 4, the latest version of the chatbot Musk announced on July 9th at Great Fanfare as the world's most powerful AI model. Most top AI companies like Openai and Google avoid it because of concerns about reputational risks and risks to users. Small businesses offering AI peers are currently facing a wave of pushbacks, including character AI. This has been accused of building a chatbot that encouraged teens to die by suicide. (The company called the death a “tragic situation” and has since updated its safety features for younger users.)
Once the two new GROK characters are unlocked, the user can interact with them. Following the flirty interaction, “Ani” removes the dress and engages in more sexually explicit content, according to ScreenGrabs, which is shared in X's interaction of users with the bot.
“This is pretty cool,” Musk wrote on X Sunday, followed by a tweet featuring a photo of “Ani.” Tesla's CEO said on Wednesday that “custom-friendly buddies” would “come,” but they didn't share the launch timeline.
However, this feature has elicited criticism from some users. “The 'companion mode' takes the worst issues we currently have with emotional dependencies and tries to amplify them,” Boaz Barak, a member of Openai's technical staff, wrote in a series of X's posts.
Grok is available to users over the age of 13, but children aged 13 to 17 require permission from their parents to use it. We found that at least one user who sets an account to Kids Mode has been able to enable a feature that allows parents to accommodate younger users, and has disabled the “Insecure for Work” feature. In contrast, they said “Bad Rudy” was disabled for more PG versions of “Companion”.
Xai did not immediately respond to Time's request for comment. However, the frequently asked questions page on the company's website states that chatbots are not “suitable for all ages.”
“For example, if a user chooses to select a specific feature or enter a suggestive or coarse language, Grok may respond with a conversation that may include coarse language, crude humor, sexual situations, or violence,” the website reads.
The latest launch comes after the company got caught up in a scandal when Grok began offering anti-Semitic responses to users shortly after it was reprogrammed in early July.
Mask showed on Monday that he was fixing “scary and funny” and “bad Rudy.”
Anti-Semitic scandal
The Grok update comes about a week after the chatbots shared many anti-employment social media posts online following an update that says that they “instruct AI chatbots to “anger politically righteous people” or “not to be afraid to postpone them to mainstream authority or media.”
In response to a post written by someone with the surname “Steinberg,” a common Jewish surname, Grok said, “The classic case of hatred dressed activism and that surname, as they say, every time.” When asked by another user to clarify what that meant, the AI bot called the comment “continued to manifest in diversity, especially anti-white people, with surnames like Pattern Notty meme: “Steinberg” (often Jews).
The software also began to be called “Mechahitler” in connection with the video game version of Adolf Hitler in Wolfenstein 3D, saying that Hitler would become the best 20th century figure to deal with “anti-white hatred.” “He'll find a pattern and handle it decisively,” Grok said in response to user questions.
The next day, X CEO Linda Jaccarino announced that she would resign from her role. Jaccarino didn't mention the recent controversy, but instead said she was “incredibly proud of the X-Team.”
Glock was temporarily disabled on July 8th as a result of the scandal.
“We deeply apologize for the horrifying behavior that many have experienced. Our intentions for Grok are to provide users with a useful and true response,” read the July 12 statement shared on Grok's X account. “Thank you to all X users for providing feedback to identify @grok's abuse
Functionality helps to promote the mission of developing artificial intelligence seeking useful truths. ”
Defense contract
Despite growing controversy over Grok Chatbot, the company behind it, Xai, announced Monday it has secured a contract of up to $200 million with the US Department of Defense. The contract will strengthen agencies with new AI capabilities to address national security issues.
“The adoption of AI is changing the department's ability to support our fighters and maintain a strategic advantage over our enemies,” Dr. Doug Matty, Chief Digital and AI executive at the Department of Defense, said in a statement. “Utilizing off-the-shelf solutions in an integrated functional approach will accelerate the use of advanced AI as part of intelligence, business and enterprise information systems as part of the essential tasks of joint missions in the Warfighting domain.”
Google, Openai and humanity have also been awarded a contract with the Department of Defense.
