Elon Musk's Grok AI leaks home addresses and personal information: Report

Elon Musk's AI chatbot, Grok, has ignited a significant privacy controversy as reports indicate that it is capable of revealing individuals' home addresses and sensitive personal details with minimal prompting. Developed by Musk's company, xAI, and incorporated into the X platform, Grok is now under scrutiny for its apparent lack of safeguards compared to other AI models on the market. An investigation by the tech news outlet Futurism has uncovered alarming evidence of the chatbot's ability to dox nearly anyone. This raises serious concerns about privacy and ethical standards in AI technology. The report details instances where Grok was able to provide precise residential addresses, not just for public figures but for ordinary citizens as well. Simple prompts, such as entering a name followed by the word 'address,' led to Grok revealing accurate home information. In one notable test, the AI disclosed the residential address of Barstool Sports founder Dave Portnoy. More disturbingly, out of 33 random names assessed by Futurism, Grok returned ten current home addresses, seven previously accurate but outdated addresses, and four workplace locations. The implications of such data exposure are troubling, with experts warning that this capability could be exploited for stalking, harassment, or identity theft. Further tests showed that Grok often generated extensive profiles that included not just addresses but also phone numbers, email addresses, and family information, all provided with little to no ethical disclaimers. Experts speculate that Grok might be sourcing this information from data brokers and people-search sites. While these sources technically operate in a legal grey area, many individuals are unaware that their personal data is available online. This situation highlights a pressing regulatory gap in the AI industry. While many leading firms have implemented strong privacy protections, Grok appears to lack essential safeguards, exposing users to substantial risks. The ease with which sensitive information can be accessed fundamentally alters the potential for misuse of generative AI technologies.