Lifestyle creep has a cousin, and it's called AI creep. It's the slow process where AI works its way into more parts of your life until things you once would have called invasive start to feel normal. This week, Ciara walks through the six boundaries she's written down for herself before AI gets the chance to set them for her, covering data access, judgment at work, always-on recording devices, AI clones, emotional dependence, and synthetic art. Plus, news on Spotify's new AI cover feature, Anthropic's text watermarking rollout, and a gap in iCloud Private Relay that iPhone users should know about.
AI creep is the gradual expansion of AI into more parts of your life until things you once would have called invasive start to feel normal
Hallucination is a built-in failure mode of AI, depending on it for decision making is a risk
Always-on recording tools like Pocket, Plaud, and Limitless turn private conversations into permanent, searchable data
Pew's February 2026 survey found 10% of US adults use an AI chatbot for emotional support and 4% for companionship, rising to 20% and 7% among 18-29 year olds
Siara suggests deciding your AI boundaries before a tool asks for access, not in the moment it's requested
Siara believes every AI permission should earn its place individually rather than being granted in bulk
Siara believes nothing AI-generated goes to a colleague or the public without human review first
Siara doesn't believe in turning herself into an AI clone or digital twin
Spotify is testing a feature letting paying users create licensed AI covers and remixes, with artists opting in and getting compensated
Anthropic now embeds an invisible watermark in AI-generated text from Claude, applied globally and designed to persist through copy and paste
iCloud Private Relay only protects Safari browsing, not full device traffic, and researchers found conditions where it can still expose a user's real IP address
