How I could have hacked any Facebook account

http://www.anandpraka.sh/2016/03/how-i-could-have-hacked-your-facebook.html

by phwd • 10 years ago

443 152 10 years ago

haser_au• 10 years ago

A great example of responsible disclosure, and the company acknowledging, fixing and rewarding the bug and finder. Great job to both Facebook and Anand.

jdcarter• 10 years ago

Good reminder here that all publicly-visible services are part of your overall attack surface, including beta sites and other things you never expect people to look at. The DROWN vulnerability from last week was similar: people disabled SSLv2 on their web servers, but not their mail servers.

Very nice find: super simple but super effective. I'm glad Facebook paid up promptly.

debacle• 10 years ago

Good on Facebook for being so quick to reward Anand and fix the issue.

cphoover2• 10 years ago

Frankly I think the amount being award by these companies is minuscule when you compare it to the amount of damage this information could have caused Facebook in the wrong hands.

ghayes• 10 years ago

I suspect most whitehat researchers would be happier to report this vulnerability and make a nice legal reward, then delve into a black hat market for selling a vulnerability. Seems pretty win-win in this case.

at-fates-hands• 10 years ago

Unless your name is Kevin Mitnick. Then you set up a business and play middlemen in selling them to anybody who wants to pony up for it.

"When we have a client that wants a zero-day vulnerability for whatever reason, we don’t ask, and in fact they wouldn’t tell us,” Mitnick tells WIRED in an interview. “Researchers find them, they sell them to us for X, we sell them to clients for Y and make the margin in between.”

http://www.wired.com/2014/09/kevin-mitnick-selling-zero-day-...

restalis4• 10 years ago

By reporting the problem to the owner you should assume that it might get fixed before you'll manage to sell it to other interested parties.

arcticfox5• 10 years ago

I agree. This was basically unrestricted access to any Facebook account. What a MASSIVE flaw.

And who knows if this was already exploited in the wild?

$15k is nothing at all compared to the scale of this issue...

wallflower• 10 years ago

$15k and a likely open FB job offer

mayneack• 10 years ago

It's not really unrestricted. Given that this is forcing the password reset, you can't silently do it, right? So anyone exploiting it knows there's a limited number of uses before people notice that their passwords are being reset by not them.

lavezzi• 10 years ago

Doesn't Facebook alert the user if it detects suspicious account login activity from an unknown location or IP?

mabbo• 10 years ago

During the fiasco that was the last white-hat hacker to report he'd hacked Facebook, I posted this:

Bug bounties are supposed to represent a high probability payoff of a lesser amount of money for finding a bug. This is in comparison to going the black hat sales root, where probability of sale might be lower, but the payoff might be higher. I can imagine one or two state actors who might pay top dollar to have keys to the kingdom to a major social network. (https://news.ycombinator.com/item?id=10756159)

Expected payout = (Probability of reward * size of reward) +/- any additional value I put into who I'm selling it to.

If I have an exploit that gives me a lot of access to Facebook (or any other large company) I have to run that formula for each entity that might pay me for it- Facebook, the NSA, the Chinese Military, and so on down the list. Facebook is offering high probability for a lower payout. The NSA might step up their game and start offering ten times as much, or make it more clear that they will indeed pay you for it (not that I have any proof they'd do such a thing).

Make no mistake, Facebook has probably done this math very carefully when choosing a bounty.

baddox2• 10 years ago

From this article I get the impression that the bounty wasn't even known before hand. It doesn't seem like a very careful consideration from Facebook, and I would expect it to be "too low" if they're choosing the amount after someone has already disclosed the bug.

ktRolster2• 10 years ago

In other words, if you wanted to game the odds, you'd start by offering it on the black market, then if there were no takers after X number of days, offer it to Facebook?

cookiemonsta• 10 years ago

where do people even go to start to sell an exploit? how does that kind of stuff work?

vbezhenar• 10 years ago

Another factor is that when you are buying on black market, you can't be sure whether you are buying real exploit or fake one. Exploit owner probably will request (irreversible) bitcoin payment, will communicate via anonymous channels and is unlikely to give out details about that exploit until he's got his money. So both sides have difficulty trusting each other. Probably solution is some trusted 3-rd party, but is there one in black market? It's hard to imagine, actually.

kzhahou• 10 years ago

Make no mistake, Facebook has probably done this math very carefully when choosing a bounty.

Having worked at several similar companies, I'm guessing it went like this:

"How much should bounty be? Is $5K good?"

"Wow, that's nothing. At least make it something meaningful like 25 grand."

"Yeah....that's gonna be tough. How about split the difference? $15K"

"Sure."

Ntrails• 10 years ago

You forget to normalise by the uncertainty of payment in the blackhat case and the probability of repercussions.

Not to mention the possibility of selling the same thing to multiple black hat buyers.

dsacco4• 10 years ago

This has been discussed many, many times on HN before. This bug would not cause Facebook much damage; in fact, Facebook and Google tend to overpay rewards for bugs for the purposes of goodwill and recruiting.

Let's examine the facts:

1. A Facebook vulnerability is dangerous to Facebook. A WordPress vulnerability is dangerous to a quarter of the internet. Facebook is not a high value target, relatively speaking.

2. A Facebook vulnerability will be patched once it is widely used. Facebook's security team is one of the strongest and most sophisticated of any company, and their processes would quickly catch this once it was used. The total impact of the bug would be negligible. You'd lose the ability to compromise accounts as soon as you tried to do it in any meaningful or lucrative way.

3. A vulnerability in Facebook might last a week before being patched, but a vulnerability in PHP will persist on the internet for years. No matter how many individual sites patch their servers, you'll still be able to pop a lonely server with social security numbers chugging along in a closet somewhere.

There really isn't much more to say about this. People claim bounties awarded by Facebook/Google/et al are undervalued every single time a bug bounty hits the front page of HN. Every single time, someone who is in the security industry patiently explains why it's not that valuable.

If someone tried to go to a blackhat group or go to the "black-market" (a shadowy, lucrative place that never seems to be very well-defined in these conversations), he would not even be able to find a seller, let alone one who would pay a lot.

What do you imagine someone would pay for this on the black-market? They'd need to profit from it. How much profit is worth their time?

Say they buy it for $20,000. Do you really think someone will derive $20,000 of profit from this before it's caught and patched by Facebook?

The only vulnerability worth $15,000 or more is one directly impacting a language, a widely used development library/framework or a widely used piece of software.

For further reading on bug bounty valuation:

https://news.ycombinator.com/item?id=7106953

https://news.ycombinator.com/item?id=9302188

https://news.ycombinator.com/item?id=9040855

https://news.ycombinator.com/item?id=9041017

https://news.ycombinator.com/item?id=8563884

comboy• 10 years ago

The point is, it's possible that if he was a black hat, he could have gotten much more for it on the black market. Even if it could only be used once or a few times.

Bounties should not only be higher than what you can get on the black market. They should be high enough to make security experts spend their time trying to win them. Given probability of finding a bug with such impact, it is just not worth the time of any decent white hat (in terms of financial gain, sure it's nice to brag about it and that I believe is the biggest motivation).

Money wise, for Facebook, it makes zero difference if they pay him that much or 5 times more.

arcticfox5• 10 years ago

What are you talking about? Facebook is not a "high value target" and "this bug would not cause Facebook much damage"?

For example, if you wanted to monetize it, I have to imagine TMZ (or someone even less scrupulous) would pay a lot of money for dumps of A-list celeb and athlete Facebook accounts.

You don't think Facebook having "The Fappening Part 2" on their hands is worth more than $15k to prevent? Or having every US Government FB account simultaneously posting ISIS propaganda?

The PR for any number of scenarios like those would be an absolute nightmare for Facebook.

agiamas• 10 years ago

yeah seriously, #2 is spot on. The moment you start brute forcing your way, network traffic gives you in and the security team will start blocking you. By the time they find out you may have found out 1,2,3.. few accounts.. This is no way worth lots of $$ in bounty, let alone the 100k job offer, that's a joke.

ttsda• 10 years ago

This would be extremely lucrative for a media outlet. You would need way less than 5 exploited high-profile celebrity accounts to surpass the $15,000 that were awarded in profit.

murukesh_s• 10 years ago

Hack into zuckerberg's id, get his card info if it's stored, and for fun change the profile pic, send the news to top blogs before letting Facebook know. How much it would cost Facebook to do damage control?

hitekker• 10 years ago

If I had a site with a billion users and said site had a vulnerability with significant repercussions on my business, BreakingBits, a "software security firm", would advise me not to meaningfully encourage outsiders to find and squash those vulnerabilities?

Uh... is this the advice you give your customers?

unabridged• 10 years ago

Say they buy it for $20,000. Do you really think someone will derive $20,000 of profit from this before it's caught and patched by Facebook?

FB is publicly traded at very high volume. You can make very large bets on its movement without being noticed. A hack of 10 celebrity facebook pages could probably drop the stock 5% in one day. You'd probably be able to make at least 20 to 1 on your money using options. The right organization could clear millions.

betolink• 10 years ago

Facebook is not a high value target, relatively speaking.

I think you got this wrong, Facebook is not the target, their users are. Which reminded me of "If a service on the Internet is free, you are the product"

ThrustVectoring• 10 years ago

Say they buy it for $20,000. Do you really think someone will derive $20,000 of profit from this before it's caught and patched by Facebook?

I can absolutely think of a situation that would make access to one person's facebook account worth way more than $20,000. Say, an unethical high-profile divorce lawyer fishing for information that would help during a multi-million dollar case. Or a political activist trying to dig up dirt on a candidate.

baby• 10 years ago

this doesn't make any sense, most vulnerability you find on products are "game over" vulnerabilities. It's common. If you would give a billion dollars to every such findings in an audit then you would be pretty quickly calling bankruptcy.

Also, I can personally live for a year on that kind of money. But that's another issue.

On the other hand, that is maybe what they would have payed for a real audit of a few days/weeks and it's not even sure the vulnerability would have been found (especially considering the size of Facebook). So yeah maybe they also deserve more.

I'm mixed.

tptacek16• 10 years ago

David is right about this, too. A $50,000 pentest of any major web property is likely to find multiple sev:hi vulnerabilities. By the logic the grandparent comment uses, those audits should cost more like $1.5MM.

pbreit• 10 years ago

I assume it made its way into the wrong hands (not by the OP) and the amount of damage appears to be zero.

TACIXAT2• 10 years ago

I do security audits and agree completely. My biggest issue is that the researcher is working for free. If nothing is found you just burnt a few weeks, if something is found the payout is usually only a couple grand.

If people enjoy doing it, or it makes sense in their currency or situation, awesome. The payouts don't get me excited though.

That's the market though. People who think it's too cheap don't play the game, people who think it's good money do.

dr_zoidberg• 10 years ago

In my country, getting paid 15k dollars would mean more than a years worth of sallary of a big infosec company. So it makes complete sense to go in for bug bounties, even if it's "low" payout, or if you spend a lot of time to get one.

Like you said, that's our market, but when working on a global scale, you have to consider pretty much everyone.

TrevorJ• 10 years ago

I think it's designed to be enough to convince would-be hackers to reveal the bug, but not enough to incentivise large amounts of people to sit around all day every day trying to break Facebook for bounties.

awqrre• 10 years ago

A bidding war between the bad and good guys would be interesting...

mcone• 10 years ago

How do companies evaluate the severity and impact of the vulnerability? I don't work in security, but it seems like this is worth more than $15,000.

cphoover2• 10 years ago

It certainly would be worth more to some people.

dsacco4• 10 years ago

Companies evaluate severity based on impact. There are different tiers of vulnerability.

A vulnerability that affects a particular website is significantly less valuable than one that affects many websites.

Companies like Google and Facebook actually overpay for vulnerabilities because 1) they're flush with cash and can, 2) it's excellent for goodwill in the industry, 3) it's an excellent recruiting tool and 4) it augments an already strong internal security program.

If you hypothetically tried to go to the black market with this vulnerability you wouldn't even find a buyer. When Facebook patches this, it's useless, and you'd have to derive more than whatever you paid for. At this point it's a betting game - do you think you can earn back $100,000 using this exploit before Facebook catches wind of it?

Conversely, vulnerabilities that are very highly valued tend to affect large numbers of websites in a format that is not easily patched. For example, many websites don't update WordPress often, which means that a vulnerability in WordPress is going to instantly get a CVE and a widespread push for awareness. Even so, it will be actionable for years.

unknownzero2• 10 years ago

Anyone know what tool he was using in the YouTube video? This stuff is super interesting.

fatlasp2• 10 years ago

Looks like Burp Suite. Sweet web proxy tool -- https://portswigger.net/burp/ Free for 14 days I think.

unknownzero2• 10 years ago

Awesome! Thanks for the link. It looks like they have a limited free forever version as well, gonna have to play with this.

msie• 10 years ago

Surprised that the well-paid developers at Facebook missed this vulnerability. Should inspire confidence on anyone who didn't get a job there. :-)

dsmithatx2• 10 years ago

This has me thinking about another possible attack. Say I don't want to hack all of Facebook or a specific account. What if I used a botnet to reset passwords and then use the six attempts randomly on each account I reset. Sure I'd only get a small percentage but, I would easily start hacking FB accounts. It's things like this that make me use 2FA as much as possible on personal data.

orionblastar3• 10 years ago

2FA is nice unless you lose your cell phone or it gets stolen. If you ever lose your job or go homeless and can't afford a cell phone then you are locked out of your accounts.

I am disabled and struggling if I miss payments I go homeless or can't pay my bills and things get shut off. For me 2FA might not work if I am down on my luck.

dsmithatx2• 10 years ago

There are ways around this for some things but, probably not with services like Facebook. For example with Amazon I use 2FA with a cellphone. As a backup I use a hardware token and an Admin account. If my cellphone gets stolen I pull the 2FA card out of my wallet.

noobie2• 10 years ago

When setting up Google Authenticator for One Time Passcodes, you are also given a seed which can be used to resteup the app from another device.

ladzoppelin• 10 years ago

Authy is really good. https://www.authy.com/

They need a Firefox extension but its allowed me to do 2FA on my personal and work accounts without fear of being totally locked out if I loose my phone.

TACIXAT2• 10 years ago

I recently bought the cheapest phone I could find for 2FA. 10$ phone and 20$ minutes. It equated for 400 texts with a time limit of 3 months. I didn't do too much research but there are probably some that don't expire. So, if you ever find yourself in a tough spot, 30$ can get you access. If you ever can't afford 30$, get in touch and I'll send you a phone.

ec109685• 10 years ago

Facebook likely makes the challenge harder if there is no history on the device you are trying to reset your password on.

thrownn• 10 years ago

On the subject of rate limiting, what is the best way to apply it across all endpoints, APIs and resources, external and internal, with minimal effort?

Usually, I see this implemented only as an afterthought, and only on endpoints deemed 'dangerous', waiting for a disaster like this to happen...

noir_lord• 10 years ago

It's a defense in depth scenario but most webservers have modules for it, Apache certainly does as I've used it not sure about nginx still not used that in production.

010a• 10 years ago

Hacker News: Where comments can be six paragraphs long and say absolutely nothing.

fabrigm• 10 years ago

avatar

dang• 10 years ago

We detached this subthread from https://news.ycombinator.com/item?id=11249116 and marked it off-topic.

nyc4life• 10 years ago

As opposed to the rest of the Internet?

beshrkayali• 10 years ago

Regardless of this being Facebook or not, but forget to throttle your API and this is what you get, some dude toying around with a tool just to poke holes in your thing, but I digress.

If in any twisted, unrealistic, straight out of Homeland scenario where anyone high profile enough would make use of this "vulnerability" and successfully create a media "splash", and assuming Facebook security team is on top of their game, this would get patched in a week tops. Keeping an eye on average number of requests coming to their API end points, especially sensitive ones, is part of their job, not a nice-to-have. I'd even think this would actually get patched within 24 hours (since the fix isn't really that difficult). I have absolutely no care or sympathy for Facebook but yeah, 15K is a lot for something like this. It's a nice catch, that's all.

technion• 10 years ago

    beta.facebook.com and mbasic.beta.facebook.com

Certificate Transparency has an interesting impact on some of the less-public servers.

https://crt.sh/?q=%25.facebook.com

A host of servers turn up in that list, which may similarly be less security tested than the main facebook.com site.

nly• 10 years ago

I'm surprised an organisation as large as Facebook don't have their own CA, and just don't issue the semi-secret stuff off the record.

evgen• 10 years ago

Running a CA is a major pain, adds auditing and other requirements that are ongoing pain, and prior to the past year or so Facebook did not issue enough certificates to make the cost worthwhile. Doing this right means adding a lot of logging and access control around a few parts of the infra stack that would manage this, so why not pay someone else to deal with the paperwork and bother? All FB certs are on the CT logs as a matter of policy, so that there are no loopholes in our current statement that if a Facebook cert is not on the CT logs you should not consider it valid; we will accept the loss of secrecy (and people launching new stuff hate it but have learned to adjust) if the end result is making it harder for someone to slide a dodgy cert into the chain.

fblp• 10 years ago

I'm surprised by the amount of certificate fragmentation these companies have. Why would they use so many different certificate types and vendors? Twitter is even worse: https://crt.sh/?q=%25.twitter.com

sandGorgon• 10 years ago

BTW - Anand is a security engineer working for Flipkart and is one if India's smartest security experts. This is not the first time he has found bugs.

http://yourstory.com/2015/10/techie-tuesdays-anand-prakash/

annnnd• 10 years ago

I would love to know if someone has exploited this bug - should be fairly easy to learn that from logs (this attack is far from stealthy). I guess FB will never tell. :)

s3arch• 10 years ago

For these individual hardworking security analysts, Facebook awarding cash prices of "any real value" is much worth than some news article reporting it as "...simple security flaw...".

http://www.zdnet.com/article/facebook-fixes-simple-security-...

moonshinefe• 10 years ago

A whole $15k? This could have cost them hundreds of thousands if not millions in lawsuits. That's a pretty crappy incentive, I'd imagine a lot less moral security researchers getting exponentially more money out of something like this by just selling the 0day.

I wonder why the reward is so low. This is literally the amount a code monkey gets paid after 3-5 months of work with minimal skills.

adam12• 10 years ago

Anyone else have trouble with that webpage? It froze my browser (Chrome).

How I could have hacked any Facebook account

Related Stories