AEM is not an archival solution! Depending on the details of the requirement you have you should consider:
* creating a content package doesn't help you in the first place. You should be able to prove that a page looked in a certain way or had certain content on it. Can you do that with a simple content package? Typically no.
* You need to have a system which is able render the content. It needs to have the correct templates, correct components, JS and CSS etc. This will change with every PROD deployment. But also your OSGI bundles might affect the rendering.
* If you think of backing up the complete instance after a deployment: Are you able to restore it? Does your OS in 2023 support to run a backed up AEM instance from 2018? Or the other way: Does AEM 6.4 support the Java version of 2023? I wouldn't think so.
* This results that you need to backup your complete VMs (including OS and Java) or at least have a recipe how to re-create them.
That's only for the rendering of a content package. But when you create that package? And how often? Do you create it on author or publish? Changing a page might affect other pages as well. And changing an asset will change all pages the asset is referenced from.
And so on ...
Then the process: All the above assumed that you were able to setup a new environment, load all your data and content into and re-create the rendered page. That's a very IT-heavy process, taking time and there's definitly risk. How often do you assume do you need this process? Once a year? Or more often? Do you test this procedure and check if it's working at all? How long is it about to take?
You see, that there's a bunch of questions to answer when you design such a process. When I was tasked with that requirement, we came up with a simpler solution. Whenever we activate a page or an asset, we determine which pages are affected from this change. And then we create a PDF of all the affected pages and drop it into an dedicated archiving solution.
This means, that we don't need to preserve backups and archives, and we don't need an restore process. Because the retrieval of such a PDF from the archive is something a business person can do on his own. Might be a bit more expensive from archiving costs, but the overall reduced complexity is worth it.
Jörg
