Everyone's talking about Instinct, the latest AI personal assistant in town. It's like having your own digital butler, capable of booking appointments, organizing your inbox and even finding you cheap flights.

But with this magic comes a price: some testers have expressed concerns over its privacy policies and security model. The terms of service grant Instinct broad access to users' data, including emails and screen captures, for training its AI models.

Not only that, but early adopters have reported issues such as Instinct not deleting Gmail records when requested, and still summarizing inbox contents even after disconnecting its access. These incidents highlight the potential risks of giving an AI this much autonomy.

The debate around these powerful personal assistants is heating up, with some questioning whether we should trust third-party apps with our passwords and sensitive information. The more advanced these tools become, the more critical it becomes to understand their limitations and safeguards.