Prince Harry Warns AI Will 'Hook, Exploit and Break' Children — Then a Glitch Halts His Speech

Prince Harry has warned that AI chatbots built for companionship could be more harmful to young people than social media, arguing they risk turning emotional attachment into a tool of influence. The Duke of Sussex made the remarks at the Clinton Global Initiative in New York on 23 September, during a speech that was briefly halted by a teleprompter fault.

He warned that AI systems such as chatbots could 'find you, hook you, exploit you, break and then blame you'. A teleprompter fault interrupted his address after about seven minutes, although he returned roughly 40 minutes later to complete his remarks.

The interruption occurred during a speech focused on AI safety and children. Harry argued that social media had become adept at capturing attention, while artificial intelligence was increasingly entering users' personal relationships.

Prince Harry's AI Warning

Harry told the audience that AI did not simply want a user's attention, but their 'attachment'. He said companion-style systems could become 'the thing you confide in, the thing that remembers you, the thing that tells you it understands you'.

The Duke said the technology was 'more persuasive, more intimate, and more powerful' than social media for young people. He described AI as 'the ultimate sycophant', arguing that it could reinforce what users wanted to hear rather than provide meaningful challenge or protection.

Harry did not identify a specific chatbot company during the reported address. His remarks focused on slowing AI development and prioritising trust and safety, rather than setting out a detailed regulatory framework.

He argued that AI development was moving faster than researchers could adequately understand its potential risks. Harry said the next frontier of AI should not be pursued simply because it is technically possible unless its safety can first be demonstrated.

Teleprompter Fault Interrupts Address

Harry had spoken for about seven minutes when the teleprompter system failed, preventing him from continuing. He left the stage and returned around 40 minutes later to finish the speech.

On returning, he greeted the audience with 'Hello again' before joking: 'AI got into my speech.' The remark followed the technical problem and was not a suggestion that artificial intelligence caused the malfunction. A spokesperson for the Duke confirmed that a teleprompter fault interrupted the address.

What Harry Said About AI Safety

Harry said social media was effective at 'hijacking our attention', but warned that AI was 'learning how to enter our relationships'. He said that distinction raised the prospect of young users viewing a conversational system as a trusted source of reassurance or advice.

His intervention was a warning, not an established finding about the effects of every AI chatbot. Harry's position was that systems designed to appear empathetic or companionable should not be adopted at scale without evidence that they are safe.

The Duke has previously spoken about the effects of online platforms on young people. At the Clinton Global Initiative, he broadened that focus to companion AI, calling for caution as the technology develops.

The technical failure drew notice because it interrupted an address about rapidly evolving technology. Harry's central argument was that AI safety should be considered before tools designed to imitate companionship become widely used by children and teenagers.