Chatbots like ChatGPT are good at scouring the web to find you that shirt or that pair of shoes you’ve asked it to hunt down. But how much of our personal information are we willing to risk in exchange for convenience?
Recently, I tried shopping with Muse for the first time, and even with my less-than-precise prompts, the bot was thorough. The search took a little while to finish, but when it did, the results were more or less accurate.
Afterward, I was prompted to choose a payment method, one of which involved connecting my payment info to Link by Stripe, a one-click checkout service. That got me thinking: If I connect my card to Link through Muse, who else might have access to that information and how is it being used?
Naturally, I then asked Muse where it stored my data, and it replied,
“I store what you share with me in your Muse account — your chat history, the memories I keep like preferences and ongoing tasks, and files in your workspace.
That comes from our conversations, things you ask me to remember, files you share, and any accounts or devices you've connected. I only pull from connected accounts for what you've asked me to do.
A few things worth knowing:
Your interactions may be logged and reviewed by Meta for safety, security, debugging, and product improvement. Chats aren't end-to-end encrypted.
I don't share your conversations or what's in your workspace with Meta's ad systems.
Conversations can contribute to developing AI at Meta — you can opt out in Settings > Data Controls.
You're in control: you can ask me to remember, update, or forget something, delete messages or files, export your info, or do a full reset in Settings > Data Controls which deletes chat history, files, and active tasks.
Meta's Privacy Policy, Muse Terms, and Muse Privacy Policy are the authoritative sources — and there's more in the Privacy Center.”
It also assured me that it doesn’t share my chat info with third parties “except when it’s needed to do what you asked — for example, if you asked me to check a store, I’ll share only what’s needed with that service to get it done.” Okay. Fine.
But a response from Muse is one thing. Should we trust the parent company behind it…you know, the one with this kind of track record?:
$725 Million Consumer Class Action (In re: Facebook, Inc. Consumer Privacy User Profile Litigation): Meta agreed to a historic $725 million settlement covering U.S. Facebook users who used the platform between May 24, 2007, and December 22, 2022. Final approval was granted in late 2023, and benefit distributions began rolling out through 2025 and 2026 via the Facebook User Privacy Settlement portal. [1, 2, 3, 4]
New Mexico State Privacy Verdict: In September 2026, a New Mexico jury found Meta guilty of willfully deceiving users and violating state consumer protection laws regarding the sale of user data to Cambridge Analytica. New Mexico is seeking billions in penalties, while Meta has contested the potential amounts. [1, 2, 3]
FTC $5 Billion Penalty (2019): The Federal Trade Commission imposed a $5 billion penalty and sweeping structural restrictions on Facebook for violating a 2012 privacy order regarding third-party data sharing. [1]
I’ll let you be the judge.
As agentic AI keeps improving, will it get harder for us consumers to resist the convenience of having these agents scour the internet to bring us whatever we need in seconds? And as we hand over more personal information, could such convenience eventually come at the cost of our privacy?
I suppose only time will tell.
