Random musings on whatever subject strikes my fancy, published every other day.

Category: InfoSec and IT Page 1 of 30

My Banks’ Websites Suck

The following is an amalgamation based on my reluctant observation of several banks’ websites.

1) The bank website is slow. When it loads it takes so long that parts of it time out… waiting for other parts. When I perform an action I am always left to wonder if it took effect. What feels like minutes later, I see that it did. I have learned to take my hand off the mouse, sit back and wait. If I can sing the Jeopardy! song four times before it settles down, there might be a problem. And of course, controls on the screen move a hundred pixels or so, a fraction of a second after I click them. This results in my having clicked something else.

2) The bank website is buggy. When I sign off, it goes back to a login screen. It’s a trap! This login screen will not work. If you log in on this screen, you’ll get inscrutable error messages that quickly disappear, and then a new login screen. The second login screen works, but leaves behind a nagging suspicion that the first login screen has stolen your credentials.

The secret turns out to be, to click on the bank’s logo on the top left, and get a fresh login screen. That one will behave properly.

3) The bank website is arbitrary. Here’s a richly detailed example: Six months or so after changing the email address I use to have Chase Bank communicate with me, I started getting this after logging in:

A UI is like a joke. If you have to explain it, it sucks.

The redacted text is my email address. On this screen, it is selectable text but cannot be edited. The two radio buttons and the blue bar show as clickable when I mouse-over, but I am afraid to click them. Nothing else on this screen is clickable. Not that there’s much else on this screen.

The URL of this screen ends with:

intercept?name=BAD_EMAIL_ADDRESS#/intercept/addOrUpdateEmailAddress/update

If I modify the URL by removing all that crap, the website obediently reverts to its normal behavior by showing me my accounts dashboard. Solid.

This all started two weeks ago, and it seems pretty plain that Chase is having opinions about my email address. Note that this is an email address I changed to back in September of last year. I have dozens of emails from Chase using this email address.

The email address has a + sign in it to allow a Chase-unique suffix, so I can be aware if they sell my email address to spammers. I suspect, though nobody at Chase will confirm, this is why they are having a fit of pique. They don’t want me to know that they were the ones who sold my email. Unlike many people, I have another recourse from this: since I have my own domain, I can create as many unique email addresses as I want. Perhaps they will find “ChaseBankSucks@mydomain.tld” less of an issue?

Since composing the above, two interesting things have happened. One, following a looooonnng call I had with their web support line this morning, Chase sent me an email telling me that I had changed my email address. I had not. But Occam’s Razor tells me, they did. The thing is, it’s still the same email I have had all this time. So they changed it to… itself? And the second interesting thing is, the last time I logged in, the black intercept screen has not reappeared. If they actually fixed something, they have decided not to communicate that fact.

4) The bank website treats me like an idiot. Because they think I haven’t the capacity to examine more than a single item of information at once, when I want to make a payment, I must click thru five screens.
The first: Verify that the payment account I am using is the same one I have been using for six years. Next Screen!
The second: Choose an amount to pay. Next Screen!
The third: Choose the date to make the payment. Next Screen!
The fourth: Now I must review all the stuff I chose on the first three screens and then, finally, actually make the payment. Next Screen! (oh, did you think we were done? You funny!)
The fifth: Review it all again, after I have committed to it on the fourth screen. Can we escape next-screen hell now? Oh, we can? kthxbai!

5) The bank website won’t use proper two-factor authentication.
Authy and YubiKey, why do I have these? Apparently to protect my Twitter account. Because when it comes to banks, they are just not in the conversation. But sometimes in banking, news of the world can leak in a little. It’s kind of muffled and distorted but they get scraps and go with them.

So banks got the idea they should have two-factor authentication. Then they misheard the part where we were saying (out here in reality) that email and SMS were so shitty for this purpose but hey, maybe better than nothing. All they heard was “…email… …SMS…” Accordingly, I have a bank that can only email my auth code. I have one that can only text-message.

I have one that can do both, and does so every single time I log in. The “Remember this device” checkbox is a placebo, because the next time I log in from the same device, it’s the same thing again. Hey: it used to do this multiple times per session – at least they’ve fixed that for now.

One still asks “security questions.” Yeah, you forgot there were even worse things than 2FA via SMS.

Then there’s the one that can only text but if I want to sign in “with the mobile app” that’s somehow cool.

I don’t know why the bank website has to be so crap. I would hope it’s programmed to be extremely resistant to hacks, but I did not think that required a bottom-of-the-barrel UX.

Trade-Offs

Risk is handled in one of three fundamental ways:

  • Mitigated: You reduce risk by enacting some countermeasure. Network attack risk is reduced when you install a good firewall. Malware risk is reduced when you roll out anti-virus software. Shoplifting loss risk is reduced when you install cameras and hire guards in your store.
  • Transferred: The risk is reduced by paying someone else to assume it. To put it simply: you buy insurance.
  • Accepted: You realize risk cannot be zeroed out even if you spend more money than you might lose to the threats. So you find your “sweet spot” and realize, some risk still remains. We call this, residual risk.

In information security, everything is trade-offs. Usually, the trade is resources for risk reduction. Finding the sweet spot is not even the hard part. The hard part is getting management to understand why the sweet spot is found where residual risk is still annoyingly non-zero.

There are other trade-offs. Every time we transact with a company, we risk some of our private information in exchange for some benefit that company offers us. We risk the disclosure of a credit card number, to gain the benefit of a new FitBit. We risk the privacy of our home address, to gain the benefit of having delivery of that FitBit to our door. We risk the privacy of our health information by putting that FitBit on our wrist, and syncing it to an app in our phone. This gains us the benefit of the aid that the FitBit provides to our exercise program.

Personally, I consider a FitBit too risky for privacy to be worth the benefits it can provide. Maybe someday FitBit will show me that the benefits can outweigh those risks.

See what the big deal is

But the ultimate case where I wonder how people are making these risk/benefit decisions comes with things like Alexa, Google Assistant, Siri, and (shudder) Facebook Portal. We’ve already seen cases where voice recordings from people’s homes have been grossly mishandled. What about the cases where they are handled “properly?” Where the “proper” handling of this data is to build a profile of you so detailed, your spouse would be surprised to learn some of it?

Maybe I’m the most digital Luddite around. But I will have none of that in my vicinity. When I am in its presence, unwillingly, I may do something like this:

Maybe that will tip someone’s risk-acceptance decision the right way. That’s me, always looking for a way to reduce that residual risk.

Ah, Facebook

Signal is at it again*. I’m really starting to love Moxie Marlinspike.

This is an ad that Signal bought on Instagram (owned by Facebook). In lieu of using the rich bouquet of data that Facebook presents to advertisers about the user, they just passed it along to the user.

The point was not lost on Facebook, who have banned Signal’s ads from the platform. It’s all pretty delicious.

Meanwhile: this.

Oh, yeah. One piece of advice. Get off Facebook.

_______________________________________

Earlier Signal fun here.

Signal’s Revenge

You may have seen a story last December that a software company called Cellebrite claimed it had cracked Signal. Signal, you should already know, is the most secure messaging app you can get. You want Signal on your phone instead of whatever garbage they shipped on there.

Well, Cellebrite was lying. All they could do was copy the encrypted files Signal keeps on the phone. Nothing at all to do with breaking the encryption.

Moxie Marlinspike, the founder of Signal, was not amused. He explained the reality of the situation. All things considered, his response was quite measured.

He was clearly biding his time. (best served cold, amirite?) Yesterday, he published this blog post about a Cellebrite data-stealing kit that mysteriously came into his possession, and how he cracked it.

Read the whole thing here. The TL,DR; is that since Cellebrite indiscriminately sucks all the data off the target phone, it’s not at all difficult to craft files that will thoroughly pwn the Windows machine on which Cellebrite is being run.

But my favorite part is at the end, when he mentions:

In completely unrelated news, upcoming versions of Signal will be periodically fetching files to place in app storage. These files are never used for anything inside Signal and never interact with Signal software or data, but they look nice, and aesthetics are important in software. Files will only be returned for accounts that have been active installs for some time already, and only probabilistically in low percentages based on phone number sharding. We have a few different versions of files that we think are aesthetically pleasing, and will iterate through those slowly over time. There is no other significance to these files.

I, for one, will be grinning about this for days.

Jeff is Listening

Do not, under any circumstances, have one of these in your home.

There is no reason whatsoever to assume that “smart devices” with microphones are not listening and analyzing everything they can hear 100% of the time.

There is no reason to assume that the mute button does anything other than turn on an LED.

There’s a reason that a 70″ LED monitor costs $4500 while a 70″ smart TV costs $399. They are making up the difference in revenue via the sales of your data. What you watch, what you eat, how much you argue, how much you have sex. What your kids talk about. If they can hear everything, they know everything.

I will have more about the state of IoT security in weeks to come. Suffice to say, things have not improved.

Page 1 of 30

Powered by WordPress & Theme by Anders Norén