Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
1Password for Claude Lets AI Log In Without Seeing Your Passwords
- Thread starter MacRumors
- Start date
- Sort by reaction score
Apple Passwords? Apple Intelligence can fix passwords automatically in the 27 betas. 🙂
The amount of people commenting without either having read the article or not understanding it is quite fascinating.
DocMultimedia
macrumors 68030
It's all about "why would I want to do this?" The article talks about how secure and safe this is but not a single reason why it's worth doing. So it's only natural that folks will fill that gap in with their own assumptions. Personally when somebody goes out of their way to say how safe and secure something is without any concrete reasons why I should buy it, then I get suspicious which then quickly turns into mockery.
You are not sharing your password with an AI agent, you are giving it access to a website that requires a login. If you want agents to do more meaningful work than set timers you need to give them permission.
canadianreader
macrumors 68000
They do. Those who care about privacy and security are still a teeny tiny minority.
That's incredible scary. While the setup sounds logical, and when going as planned, helpful where the app doesn't see or store the logins, and the user has to approve. But of course the issue is when there is a bug and it doesn't go as planned, where it gets it without approval, or somehow does get access to the password and or saved somewhere, or even if works as planned but user gets desensitized and just grants permission without thinking or allows permission but once in the account Claude/ChatGPT takes an undesirable action.
But at least this is being down with 1Password, since if they don't others will and then people will be installing far less secure plugins or scripts to do it, so while it's presumably the feature, definitely wouldn't want to be the first ginuea pig on this one.
But at least this is being down with 1Password, since if they don't others will and then people will be installing far less secure plugins or scripts to do it, so while it's presumably the feature, definitely wouldn't want to be the first ginuea pig on this one.
xaqt93
macrumors 6502a
I don't see an issue with this. The AI is not seeing your passwords and you can revoke access at anytime. Plus, you are already giving your passwords and information to 1Password...so...what is the problem if you give access to it? The AI agent will only ever do what you tell it to do. And the cool part is, you can turn it off whenever you want! Its not like you have to use this feature.
I personally used Apples AI agent on the new betas to updates my passwords and it worked great! I don't see what the problem is with this at all.
I personally used Apples AI agent on the new betas to updates my passwords and it worked great! I don't see what the problem is with this at all.
This article assumes that websites will allow agentic AI to access on your behalf. Current legal precedent says no they won't. Amazon, LinkedIn, just to name a couple.
What I meant is that many posters seem to misunderstand what "this" is in this case.
This is not about giving an AI access to your personal passwords. It's not even necessarily about giving an AI access to any password. It's about providing a possibly safer way to use AI and have it login to a website exactly without the need for it to know the password. And for some use cases, logging in with a password is a hard requirement.
Could you also use that to have the AI log in to the MacRumors forum for you? Sure.
Should you do that? Probably n... \xFF\xFE\x00\x1B□▪▫◊◊◊¤¤^@^H^[[H^[[2J
Last edited:
And what happens if secrets were exposed? Does 1Password then let you go back in time and undo autofill?
I mean, it’s great that 1P will check to make sure nothing has leaked, but that needs to be done on a far more proactive basis, not after the fact.
This is like locking the door to your house, and then getting a notification that the door didn’t lock properly and your place was robbed.
AdmiralKirk
macrumors member
There is nothing actually wrong with this.
If I want an AI agent to log into a website and do something for me, it needs the username and password. This way I get to see what it’s doing, and specifically approve that password only.
People get crazy with AI stories. The risk here is that the agent will do something unintended with the website once it’s logged in, not that it will steal all my passwords, because the way this is set up, it can’t.
If I want an AI agent to log into a website and do something for me, it needs the username and password. This way I get to see what it’s doing, and specifically approve that password only.
People get crazy with AI stories. The risk here is that the agent will do something unintended with the website once it’s logged in, not that it will steal all my passwords, because the way this is set up, it can’t.
chrisninjabay
macrumors newbie
Oh hell no!
What’s up with these AI-Sheep and their willingness to unbrain themselves?
What’s up with these AI-Sheep and their willingness to unbrain themselves?
BarredOwl
macrumors 6502
turbineseaplane
macrumors Penryn
Captain Kirk was up a few posts.
I’m not sure where Spock stands on this.
He might find it very logical.
Be careful if Lore starts recommending it though.
surferfb
macrumors 68030
I’m honestly a little surprised at the negative reaction here. If someone is using agentic AI tools, providing them a way to share usernames and passwords with the agent without exposing the usernames and passwords to the agent is a good thing.
turbineseaplane
macrumors Penryn
This story in isolation is of course just fine and good!
For me I'm just concerned with where 1P will go on the where/how/what re: AI
That's not new for me though, as I love 1P but have become increasingly concerned about directionality with the ownership changes.
Seeing their ADs at F1 (and elsewhere in high profile spots) hasn't helped, as it reminds me of the need (I'm sure) for "number to go up" incessantly.
Hear is a scenario where an agent could be useful. --- Insure that my bank account has enough money for my upcoming auto-payment. If not transfer money from my savings account and if my savings needs more money notify me. In this case, the agent needs my passwords for my bank and credit card. This is much more useful than having an agent turn in some lights.
