So yes, it was bad form for OP to not establish billing alerts, but you can't in any way say that AWS is transparent about most aspects of their service tiers.
AWS can certainly do a better job in telling me how to optimise my AWS hardware and expenses
In OPs case, if you search in Google, you will find this AWS hacking happen a lot with many users including me and AWS support was extremely kind to me to give additional credits to offset that expense.
It would be trivial for AWS to force you to setup spending limits when setting up an account, one for an alert, one for a hard lock.
AWS billing is terrible too, even now I'm getting charged a few dollars on my personal account for who knows what, I've cancelled everything I can find. It's like whack-a-mole everytime you spin something up.
In the end that's basically theft by AWS but you can't make too much noise or your account gets cancelled.
Anyway I was talking more about cloudwatch monitoring, including tracing all API calls. The billing is secondary: someone has been in his account and might have exfiltrated everything.
I myself have never created an instance, I set up the account then gave access to the devs. I only log into the account to provide new access to devs that need it and none of them are full time.
Ultimately the responsibility lies with me, but I would disagree that my dev team is clueless. Rather they're working on development, not watching what the servers are doing on a daily basis so I think that's a bit unfair.
I assume you revoke access later, but I doubt you audit anything that they may have done (like create keys that outlive their access) in the account or that any of it is version controlled or traceable.
And you’re surprised you’re in this situation?
And fair enough, you pay your contractors to do a specific job. None of them are going to point out that the way you’re managing your infrastructure is pretty slow and inefficient, or that perhaps there’s a better way to do any of what you’re doing on AWS that is cheaper, faster, more secure and that might give you a far quicker iteration time with the added advantage that you won’t fall apart with a surprise bill like this. They, after all, are working on “development”.
It’s not on AWS to flag anything. They give you the tools to comprehensively monitor your account, but you choose not to use them. Cloudwatch is also “standard”. Datadog has a free tier, I’d suggest checking that out because you don’t seem to have any infrastructure monitoring?
I’d honestly hire some people that have experience in this area, because your “dev team” sounds clueless.