Instinct's founder says the viral AI agent's hallucinations can spit out what looks like a stranger's data
The founder of the AI assistant Instinct, Noah Shinn, claims that reports of the tool leaking a stranger's personal data were actually the result of an AI hallucination. Shinn maintains that user data remains partitioned and secure despite the AI fabricating specific details during a user interaction.
Why it matters
The incident highlights the ongoing challenges of trust and reliability in generative AI, particularly regarding privacy and the tendency of models to fabricate information.
Personal AI agents like Instinct are taking over the tech world. Dado Ruvic/REUTERS The founder of a viral AI assistant said it can hallucinate and appear to send you a stranger's details. Noah Shinn said Instinct did this by fabricating a proper noun and then amplifying it with its reasoning. It came after a user said on X that Instinct told them a photo had "crossed into" their conversation. If your personal AI assistant sends you what looks like intimate details about someone else's life, chances are it made up the whole thing. At least, that's what Instinct's creator, Noah Shinn, said happened when a user of the viral AI tool said an agent admitted to leaking someone else's conversation into their chat.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in