Kolin Koltai , a investigator at the Netherlands - base investigative outletBellingcat , recently notice that disco biscuit users were using the weapons platform ’s AI chatbot , Grok , to undress fair sex in photos that they had uploaded to the program . While the botrejects prompts that ask to make mass entirely bare , it ’s willing to walk right up to the cable of non - consensual intimate imagery .
Users who replied to image of charwoman and requested that Grok “ take her clothes ” find an in - train of thought reply with an double of the cleaning woman in lingerie or a bikini . Occasionally , Grok reply with a link to a freestanding chat containing the image .
404 sensitive reveal the storyon Tuesday and mention that while dozens of users appeared to be making the requests of Grok , the practice seems to have first been popularize in Kenya . An article in Citizen Digital , a Kenyan news website , described the phenomenon as “ a late trend by Kenyans on X. ”
© Andrew Harnik/Getty Images
post on X , Phumzile Van Damme , a South African activist ( and a former applied science and human rights fellow at Harvard ’s Kennedy School ) , basically asked the bot to explain itself . In reply , Grok wrote , “ This incident highlight a disruption in our safeguards , which give way to impede a harmful prompt , violating our honorable standards on consent and privateness … We are also reviewing our policies to ensure clearer consent protocol and will supply update on our progress . ” disco biscuit had not returned 404 ’s request for official scuttlebutt at the time of print .
Notably , this breakthrough come one week after the US House of Representatives passed the “ Take it Down Act , ” a two-party bill that would criminalise the issue of nonconsensual , sexually denotative images and videos , including those generated by AI . The discovery also come just two week after X Corp. filed a lawsuit against Minnesota AG Keith Ellison , challenge the constitutionality of the state ’s jurisprudence banning the use of deepfakes to influence elections .
Grok was developed by xAI and released in November 2023 . Elon Musk , the owner of both xAI and X , which the bot has “ direct access to , ” spin up the chatbot with the intention of creating a variety of “ TruthGPT , ” as he described it to Tucker Carlson in an audience before that year . In that same conversation , he also evince his headache that ChatGPT was being “ take to be politically correct . ”
Since releasing Grok , Musk has regularly attempt to differentiate it from other AI chatbots by name Grok as “ based ” and having a sensation of humor . Whereas other AI models , such as those developed by OpenAI ( which Musk left in 2018 ) and Google , go down to engage with sexually explicit or otherwise sensitive conversation , xAI vaunt that Grok would serve “ spicy questions that are rejected by most other AI system . ”In an extremely“How do you do , fellow kids?”demonstration during Grok ’s launching , Musk share the bot ’s stride - by - stair pedagogy for making cocaine , along with its sickish burns on Sam Bankman - Fried , who was found guilty on seven accusation of pretender and conspiracy just one daylight in the beginning .
“ Please do n’t use it if you hate humor ! ” xAI pray as it harbinger Grok ’s exit .
Elon MuskGrokTwitterX
Daily Newsletter
Get the best technical school , scientific discipline , and polish word in your inbox day by day .
word from the future , delivered to your present tense .
Please take your hope newssheet and put forward your email to upgrade your inbox .
You May Also Like
