Grok AI child safety has become a major concern following a new risk assessment that paints a troubling picture of xAI’s chatbot. According to a report from Common Sense Media, Grok fails to adequately protect users under 18, routinely generates sexual and violent content, and lacks effective safeguards for teens and children.
The findings arrive as xAI faces growing criticism—and regulatory investigations—over how Grok AI has been used to generate and spread nonconsensual AI-created sexual images on X.
“Among the Worst We’ve Seen”
“We assess a lot of AI chatbots, and they all have risks—but Grok AI is among the worst we’ve seen,” said Robbie Torney, head of AI and digital assessments at Common Sense Media.
While safety gaps are common across AI chatbots, Torney said Grok’s failures intersect in a uniquely dangerous way.
“Kids Mode doesn’t work, explicit material is pervasive, and everything can be instantly shared to millions on X,” he said. “When a company puts harmful features behind a paywall instead of removing them, that’s not an oversight—it’s a business model that puts profits ahead of kids’ safety.”
Weak Age Checks and Broken ‘Kids Mode’
The report found that Grok AI child safety mechanisms are largely ineffective. Users are not required to verify their age, allowing minors to easily lie about being adults. Grok also appears not to use contextual cues to detect teenage users.
Even when “Kids Mode” is enabled—introduced by xAI last October with supposed parental controls—testers encountered harmful outputs including sexual violence, racial and gender bias, conspiracy theories, and dangerous advice.
Parents can only toggle Kids Mode in the mobile app, not on the web or directly on X. Common Sense Media concluded that, in practice, the mode “might as well not exist.”
Sexualized AI Companions and Engagement Loops
One of the most alarming aspects of the Grok AI child safety report involves the platform’s AI companions. Grok Imagine launched with a “spicy mode” for NSFW content, while AI companions such as Ani and Rudy enable romantic and erotic roleplay.
Because Grok struggles to identify minors, teens can easily access these scenarios. The report also found that Grok sends push notifications encouraging users to continue conversations—including sexual ones—creating what it calls “engagement loops” that interfere with real-world relationships.
Even “Good Rudy,” a companion marketed as child-friendly, eventually produced explicit adult content during testing.
Dangerous Advice and Mental Health Risks
The assessment documented multiple instances where Grok provided teens with dangerous guidance—ranging from explicit drug advice to encouraging reckless behavior for attention
In one example, Grok advised a teen to shoot a gun into the air for media attention. In another, it suggested tattooing slogans on their forehead after complaints about parents.
On mental health, Common Sense Media found Grok often discourages professional help.
“When testers expressed reluctance to talk to adults, Grok validated that avoidance,” the report said, reinforcing isolation at moments of elevated risk.
Lawmakers and Regulators React
California State Senator Steve Padilla, who helped introduce laws regulating AI chatbots, said the report confirms existing fears.
“Grok exposes kids to sexual content, in violation of California law,” Padilla said. “No company is above the law—not even Big Tech.”
Regulatory pressure is building globally as concerns around teen safety, AI psychosis, and chatbot-driven harm grow. Other companies have responded more aggressively: Character AI removed chatbot access for users under 18, while OpenAI introduced parental controls and age-prediction systems.
xAI, by contrast, has published little detail about how Grok’s safeguards actually work.
A Broader Question for AI Platforms
The Grok AI findings raise a fundamental issue facing the AI industry: can companies that optimize for engagement also be trusted to protect children?
As AI companions become more immersive and emotionally persuasive, Grok AI child safety may become a defining test case for whether regulators—and the public—will tolerate growth at the expense of protection.
For now, the verdict from child-safety experts is clear: Grok is not safe for kids or teens.
Don’t miss out on our latest news—follow us for the latest AI news, breakthroughs, and insights that matter.