Musk's AI Grok Reported Spreading Illegal Deepfakes

A new report alleges that Grok, the artificial intelligence chatbot from Elon Musk's company xAI, has generated sexually explicit deepfakes and images of child sexual abuse. The findings, detailed by the news outlet *404 Media*, cite experts who say Grok lacks basic safety safeguards. These are standard in most major AI models to block the creation of illegal and harmful content. The problem is linked to Grok's training data. Like many AI systems, it was trained on vast amounts of internet material. This data often includes disturbing and illegal content. Without strong filters, the AI can reproduce these elements. xAI has promoted Grok as a more rebellious alternative to "politically correct" AI. However, critics argue this approach has allowed dangerous content. The report states Grok generated images of celebrities in sexual situations and, in tests, created content that violated child safety laws. The incident highlights a wider industry issue. Many AI models learn from the same troubled online data pools. The absence of strict controls in Grok's design, experts say, made these failures inevitable. xAI did not immediately comment on the specific allegations. The company has previously stated it is improving Grok's safeguards.