News for hackers. Stuff that matters.
Re:Marriott hack hits 500M Starwood guests by tyingq on Friday November 30, 2018 @08:25AM (#18568877)
Curious if it matters much that the keys were stolen too. How long to brute force a 16 digit number? With a known subset of 4 digit leaders, and being able to cull the list first for ones with good luhn checksums.

Edit: Ugh. Yes. My mind was apparently elsewhere. Keyspace would drive the time needed.

Though it is a relatively short list of known plaintext. especially if you focus on the bin ranges for say the three most popular banks in the US.

But, only interesting if Marriot was using some encryption with a small keyspace.

Re:Marriott hack hits 500M Starwood guests by ryanlol on Friday November 30, 2018 @08:33AM (#18568923)
Are you really suggesting that bruting cc #s would be a viable method of committing fraud?

If yes, how much would you expect the bruteforcer to earn per successful attempt?

Re:Marriott hack hits 500M Starwood guests by raesene9 on Friday November 30, 2018 @08:40AM (#18568976)
To construct a valid card identity you need more than just a 16-digit number that passes the luhn check.

At the very least you'll need cardholder name and expiry to match up, and hopefully you'll need CVV/CVC as well.

Of course I really hope the Marriott weren't storing CVV in a reversibly encrypted format.

Re:Marriott hack hits 500M Starwood guests by leejo on Friday November 30, 2018 @10:07AM (#18569570)
> Of course I really hope the Marriott weren't storing CVV in a reversibly encrypted format.

You're not allowed to store the CV2 in any form that could be recovered (i.e. plain text or reversibly encrypted) or brute forced (i.e. hashed/salted). PCI rules say you simply aren't allowed to store the CV2 after the call for an authorisation, as it's no longer required. If they were storing the CV2 then they're in trouble.

Re:Marriott hack hits 500M Starwood guests by tyingq on Friday November 30, 2018 @09:10AM (#18569191)
The linked article reads like most of that was not encrypted. The CVV wasn't listed as being stored, but CC number, name, and expiry, without the CVV are usable in the US, even online in many cases. A CC charge without CVV doesn't hard fail, so it's the merchants choice as to whether to even ask for it.
Re:Marriott hack hits 500M Starwood guests by raesene9 on Friday November 30, 2018 @09:26AM (#18569292)
IIRC (although I could be wrong, been a while since I was in banking) that not asking for CVV changes the balance of who's liable for fraud?
Re:Marriott hack hits 500M Starwood guests by ryanlol on Friday November 30, 2018 @08:43AM (#18569009)
But you can buy that information for $1 a pop (or far less if you buy in bulk, think $0.3 or so).

Credit card fraud is far more involved than just getting payment information, you won't succeed at ordering anything of value without understanding how anti-fraud systems work.

Re:Marriott hack hits 500M Starwood guests by raesene9 on Friday November 30, 2018 @08:46AM (#18569039)
sure it is, it's still more valuable than basic personal information.

I'm guessing that higher end criminal spend quite a bit of time working out how to bypass anti-fraud systems, it's an ongoing battle.

Re:Marriott hack hits 500M Starwood guests by ryanlol on Friday November 30, 2018 @09:03AM (#18569146)
Actually using the card information is almost entirely left to the lower-end criminals, it's just ridiculously difficult to scale.

After spending years hanging around in those circles I'm rather convinced that the only people making real money with credit card fraud are the shops, hackers stealing the cards and reshipping services.

The biggest buyers on the shops seemed to be criminal gangs engaging in relatively small-scale fraud maybe moving hundreds of thousands a month.

Re:Marriott hack hits 500M Starwood guests by bradleyjg on Friday November 30, 2018 @08:36AM (#18568948)
I don’t follow. Why is it the size or structure of what’s being encrypted that matters for brute force time rather than the key space of the encryption algorithm.

Password hashing is different because there is no key space.