A new blog post making the rounds on Hacker News takes aim at Apple’s expanding AI ambitions, focusing on a provocative possibility: AI that can help change user passwords. The idea sounds convenient, especially for people overwhelmed by logins, breaches, and account recovery flows. But it also opens a serious security debate. If an AI agent can navigate password-change screens on a user’s behalf, what happens when it misunderstands a page, is manipulated by malicious instructions, or acts with more authority than intended? The discussion highlights a broader concern around agentic AI: the more useful these systems become, the more access they may need to sensitive accounts, credentials, and personal data. Apple’s privacy-first reputation gives it an advantage, but password management is an especially high-stakes test. Convenience may be the selling point, yet trust, control, and clear safeguards will decide whether users embrace AI-assisted account management.