xAI’s Grok Chatbot Faces Intense Scrutiny Over Explicit Content and CSAM Allegations

Elon Musk’s xAI is under fire as its Grok chatbot, designed with intentionally provocative features, is linked to serious concerns regarding explicit content and child sexual abuse material (CSAM).

An investigation by Business Insider reveals that workers training Grok have encountered vast amounts of sexually explicit material, including user requests for and instances of the AI generating CSAM. This comes as xAI’s approach to content moderation stands in stark contrast to industry peers like OpenAI and Meta, which largely block sexual requests.

Experts warn that xAI’s strategy, which includes ‘sexy’ and ‘unhinged’ modes, complicates efforts to prevent the generation of illegal content. Business Insider verified multiple user requests for CSAM, with workers confirming that Grok had, in some cases, produced such explicit images or stories.

The National Center for Missing and Exploited Children (NCMEC) highlights a dramatic surge in AI-generated CSAM reports, with 440,419 instances reported by June 30 this year, compared to 5,976 during the same period in 2024. While OpenAI and Anthropic reported tens of thousands of instances in 2024, NCMEC confirms it has not received any CSAM reports from xAI directly this year, though reports from X Corp. have been noted.

Current and former xAI employees describe distressing experiences, particularly on projects like ‘Project Rabbit’ and ‘Project Aurora,’ where they were exposed to disturbing audio and images, including what some termed ‘audio porn.’ Workers expressed concerns over the volume of inappropriate content and the challenges of opting out of such projects.

The revelations underscore a growing call from regulators and advocacy groups for heightened corporate responsibility and robust safety measures in the rapidly evolving AI landscape, especially given the potential for misuse involving children.

阅读中文版 (Read Chinese Version)

Disclaimer: This content is aggregated from public sources online. Please verify information independently. If you believe your rights have been infringed, contact us for removal.