Grimes Warns AI Is the 'Biggest Imminent Threat' to Children — Urges Caution on Outsourcing Thought

Grimes Says Unrestricted AI Use Could Harm Developing Minds

Since Elon Musk launched xAI in March 2023 and introduced the chatbot Grok, he has been a central figure in public discussions about artificial intelligence. As Musk continues to shape the public debate, his former partner Grimes has begun to voice stark concerns about how AI may affect children and young people.

Grimes — who dated Musk from 2018 to 2022 and shares three children with him — raised those concerns during an October 2025 appearance on Joshua Citarella's "Doomscroll Podcast," a conversation later published on YouTube. She opened the episode by warning about the dangers of "outsourcing thinking" to AI tools for people whose brains are still developing.

"To me, the biggest imminent threat is just outsourcing thinking," Grimes said. "I do not think when you have a developing brain, probably before 24, 25, you should be using a tool that can write for you. I think that's an incredibly dangerous thing."

Her concerns, she said, are informed by emerging research and media summaries that suggest heavy reliance on AI can change how people think and perform on cognitive tasks. Grimes cited reports indicating that adults who lean on models like ChatGPT may show declines in certain neural, linguistic and behavioral measures.

It is important to be precise about the research: scientists have not established that AI use causes literal brain-cell loss or irreversible "brain atrophy." A June 2025 Time article summarized an MIT study that divided participants into groups to answer SAT-style questions and found the cohort allowed regular ChatGPT access underperformed on several measures and tended to decline further the longer they depended on it. Those results point to potential harms from overreliance — such as reduced practice of critical thinking and problem-solving — but do not prove permanent neurological damage.

Grimes argued that, given the uncertainty, unfettered AI access for children amounts to experimenting on developing minds. "ChatGPT without a wrapper is, in my opinion, probably not safe for a developing brain," she said, urging more data and safety measures. She compared the situation to earlier public-health mistakes: "Smoking, McDonald's, and now we're going to do it with social media."

Still, Grimes said she recognizes AI's promise. She told the podcast she is intrigued by the prospect of superintelligence and its potential benefits, but that parenthood has made her more cautious about risks to children.

Musk's Longstanding Warnings — and Optimism

Elon Musk himself has a lengthy record of both warning about and promoting AI. In a March 2018 interview with CNBC he warned AI could be "far more dangerous than nukes" and questioned the lack of regulatory oversight. At the same time, he has also described optimistic scenarios in which advanced AI and automation could drastically reduce the need for traditional work — potentially enabling a higher quality of life for many people, as he described to CNN in May 2024.

The conversation around AI's risks and benefits — especially for children — remains unresolved. Researchers, technologists, educators and parents continue to debate which safeguards, usage limits and educational strategies are needed while better evidence is collected.

Originally reported by Men's Journal on Nov 7, 2025.