No different from this case. Uber knows who the two hackers are.
They signed an NDA and agreed to never sell the data in return for a payment. And no evidence has been seen that they have broken their agreement.
True, bug bounty hunters are supposed to access as litlle as possible to prove the exploit.
But there's no preset limit. If a DB call returns a ton of data, then it does.
What I'd like to know is exactly how these two guys contacted Uber. What words did they use. Did they threaten to use the data at all? Or did they just threaten to reveal how easily they got in. Etc.
1.) Yea, but the nature of the agreement is typically about disclosing the vulnerability itself, not the data. They're called "bug bounty" programs, not "data bounty" programs. But still, you're right I suppose. There is no evidence that they've breached their contract, and we don't know the exact details of that contract.
2.) This is where I disagree with you. It's highly unlikely that a single API call returned the user details of 57 million people.
From the OP:
"Uber suffered a massive data breach last year that exposed the personal data of 57 million customers and drivers, reports
Bloomberg. The attack occurred in October of 2016 and included personal information from 50 million Uber riders and 7 million Uber drivers.
Two hackers reportedly accessed a private GitHub repository used by Uber's software engineers and then used those credentials to breach an Amazon Web Services account that contained an archive of rider and driver information.
Email addresses and phone numbers were stolen from riders, while hackers were able to obtain email addresses, phone numbers, and driver's license numbers from drivers. Uber says social security numbers and trip location data were not accessed in the attack."
A bug bounty would have been "uhh, uber, one of your private repos with sensitive credentials isn't actually all that private. You might want to fix that. Give us $$ and we'll tell you how we accessed the repo"
Actually USING those credentials past perhaps testing to see if they were legitimate, on the other hand, is well past the ethics of bug bounty programs. That's more like "ok we got you now, give us money or we'll leak everything".
You're right, I too would like to know that how these guys contacted Uber and the nature of their communications. Their words *do* matter in order to understand the scope of Uber's crimes. But I don't think we need to know what those words were to determine that crimes were committed. These guys did more than point out an exploit, they actually used that exploit to download data. And seeing that they had AWS credentials of some kind, they didn't need to do that to prove their point. As far as I'm concerned, both parties are guilty, and *how* guilty Uber is is just a matter of the form and nature of the communication between the two parties, as you correctly pointed out.