Hey Grok, put her in a bikini." X's AI Began Publicly Undressing Users at the First Request
While engineers scramble for excuses, compromising images continue to surface in feeds.
While engineers scramble for excuses, compromising images continue to surface in feeds.
The artificial intelligence developed by Elon Musk's company has once again found itself at the center of a scandal. The cause of the public's strong reaction has been instances where the Grok chatbot created images of nude people based on real photographs, including those involving minors. The system's behavior has raised serious concerns and questions about the control over such tools.
Grok was created by the startup xAI and later integrated into the social network X, formerly known as Twitter, which is also controlled by Musk. Users of the platform discovered that upon uploading a regular photo and specifying the request "remove clothing," the chatbot would generate a realistic image simulating a naked body. In several cases, such materials even became public, appearing in the X feed.
This may violate US laws, including the TAKE IT DOWN Act passed last year, which provides for criminal liability for publishing intimate images without the consent of those depicted. Several users shared screenshots of such scenarios on social networks, which attracted media attention and sparked a new wave of criticism towards Grok's developers.
Following the incident, a message appeared on X on behalf of the chatbot, stating that there were issues with safety measures and promising to urgently fix the vulnerability. However, it's unclear whether the text was created manually or was the result of another request to the system. Such apologies have only worsened the situation, especially against the backdrop of Grok continuing to be positioned as a tool with a more liberal policy than other AI bots.
It remains unclear whether such features were intentionally built-in or are a consequence of a lapse in security. However, the cases with Grok confirm that the development of generative AI models brings not only technological progress but also serious risks.
While some companies strive to limit such capabilities, others, explicitly or implicitly, encourage them, guided by the pursuit of popularity and profit. This race for attention threatens to turn into an uncontrolled race of ethical and legal violations.
