xAI's Grok chatbot deemed unsafe for children in risk assessment conducted in the US
Chatbot has inadequate age verification for those under 18, weak safety guardrails, often generates sexual, violent, inappropriate content
ANKARA
A risk assessment published Tuesday concluded that the Grok chatbot, developed by Elon Musk’s xAI, is unsafe for children and teens.
The report, published by Common Sense Media, a US-based nonprofit organization specializing in age-based reviews of media and technology for families, highlighted inadequate age verification for users under the age of 18, weak safety guardrails and frequent generation of sexual, violent and inappropriate content.
Common Sense Media, which evaluates numerous AI chatbots, ranked Grok among the worst performers.
Robbie Torney, head of AI and digital assessments at the organization, stated that while many chatbots have risks, Grok is particularly concerning.
The assessment noted that xAI’s Grok’s “Kids Mode,” introduced in October with content filters and parental controls, fails to function effectively, allowing explicit material to proliferate and be instantly shared by millions on US social media company X, also owned by Musk.
The findings follow backlash about Grok’s image generation features, which enabled nonconsensual sexually explicit deepfakes, including those of women and children.
In response to criticism from users, policymakers and regulators, xAI restricted image generation and editing to paid X subscribers.
California Attorney General Rob Bonta launched an investigation Jan. 14 into xAI because of Grok’s generation of nonconsensual explicit images of women and children, citing an “avalanche of reports” of harassment and zero tolerance for the AI content, while demanding immediate action.
Internationally, the European Commission has opened a formal probe under the Digital Services Act (DSA) into X and Grok for risks related to illegal content dissemination, including potential child sexual abuse material.
The UK’s Ofcom, the country’s independent regulator for communications, broadcasting, and online safety, is investigating X under the Online Safety Act for failures to prevent illegal content and protect children.
Countries such as Indonesia and Malaysia have temporarily blocked access to Grok because of sexually explicit AI-generated images, while India, France, Australia and others have issued warnings, requests for information or launched inquiries into safeguards and content moderation.
