Ask HN: Is Zero Trust Architecture Overkill?
4 comments
is the debug service the same as the data service?
The debug service and the data are the same application - however there is a 3rd party integration while processing the scan. We are using an open source LLM to process the source code and conduct the scan via DeepInfra. This is based upon the understanding that everything sent to them is covered by a zero data retention agreement. Is that what you meant?
there might be different interpretations of zero trust here...for me zero trust means not sharing source code
My partner and I have built Dev In A Box - a debugging tool that uses simulations to pull out both bugs & security vulnerabilities with ~70% accuracy. It works for both finding the root cause of reported issues and doing more general debugging / tech debt removal sessions.
Naturally, the tool has to handle people’s code, so we wanted to make sure our security was air tight.
We’ve designed our architecture with this in mind - essentially, each new project lives in its own database. Making the contract to access the data much simpler to enforce and making a tenant breach via a SQL / coding mistake impossible.
The other goal of this is enabling a zero trust encryption system - allowing the user to prevent us from accessing their data at rest. However I’m not sure what is the best / most user friendly way to do this is. I’d appreciate any feedback on this concept:
The idea is a user could set a “master password” that only they know (and we don’t) that would serve as the encryption key. Meaning their database can only be unencrypted at run time with that “master password” - similar to how password managers like LastPass work.
This would mean we, as the SaaS company, won’t be able to access their data. Potentially giving users more peace of mind and an additional layer of security.
I think it sounds good in theory - as someone who refuses to use GitHub because I don’t trust Microsoft as far as I can throw them, it definitely fulfills my neurosis. But before we build it, I wanted to run it by all of you to ask if it’s really worth it.
There are some drawbacks, of course. Like:
- Having to input a master password every time you want to access your project
- Managing that password across larger teams (if 50 people know the password, is it really providing any security?)
- If the password is lost, there’s no way to recover the project. It’s gone for good.
So my questions to you are:
- Do the pros out weigh the cons? Is this something that would make you feel better about using a code scanner?
- Is this level of security / privacy important to you?
- Is there a better UX for something like this?
- (For the folks who are knowledgeable in cryptography) Is there a better way of going about this? Do you know of a better method for managing the data so that the provider can’t access it at run time?
For more context you can see our website: dev-in-a-box.com
Thank you in advance. All opinions welcome!