8 ms·

I long for the future where I can safely assume my passwords are stolen

96 points by jenandre 13y ago

evv 13y ago
This is a clever security enhancement and wouldn't be hard for anyone to implement. Login data could also be sent over email to a specified address. We need an open standard for this.
Delphie 13y ago
Am I the only one who thinks that medium.com is abusing the HN ranking system?
- rgbrenner 13y ago
  in what way? they have 2 articles on the front page, 0 on the second page, at this moment.
  - pi18n 13y ago
    They've had quite a few posts on the front page in the last week or so.
- drivebyacct2 13y ago
  That's like saying wordpress.com or blogspot.com are abusing the ranking system...
  - Sprint 13y ago
    Those are not curated while medium.com is "creating content is limited to a small list [of users]".
- jacquesm 13y ago
  Yes, definitely. I noted as much yesterday in another comment.
eksith 13y ago
Unfortunately, passwords are to account security what PHP is to web development. It may not be the best, it's not the fastest, it even may not be the most secure. But it's ubiquitous, it's free and if done right, it's an OK way to get the job done precluding more complex solutions that aren't as practical for most of us. I think I posted this elsewhere or maybe here, but I got around the password remembering problem by keeping one master file of all passwords/usernames/emails etc... that I keep GPG encrypted ( specifically http://www.gpg4win.org/about.html http://www.gpg4win.org/about.html ). All passwords in the master file are randomly generated. That is, of course, the account I'm creating is a throwaway at Paul's Peanut Brittle shop. Interestingly, paulspeanutbrittle.com isn't taken.
hooande 13y ago
This sounds like a good idea. I think it gets one thing wrong: Hacking living social was not about gaining access to anyone's living social account. It was about getting passwords that were reused on sites that actually matter (banking, etc). A notification that someone has logged into your livingsocial|facebook|twitter|linkedin is only valuable if it compels you to immediately change all of your other passwords. Which might be a decent idea for a version 2 of this hypothetical software...automatically change all of my passwords whenever one of my accounts is compromised. The sad thing is that people reuse their passwords so often. Just remembering two passwords, one for important stuff and one for social stuff, would go a long way toward making this a non-problem and taking a security burden off of a lot of startups.
glurgh 13y ago
Wouldn't an attacker who's gained enough access to a system to get db dumps or modify serverside code to the point where they can capture incoming passwords simply turn off the audit calls? Wouldn't an attacker who's gained access to a secondary account through a password derived from a db dump change the audit URL? Beside that, knowing an account is compromised is better than nothing but often not particularly useful - see all the horror stories by people who've had their gmail account compromised. They tend to find out very quickly the account has been taken over, recovery is still difficult and the loss of data/time/neurons is often significant.
- chacham15 13y ago
  A user isnt only vulnerable when the db gets hacked. People have other ways at getting at passwords (sometimes they guess). The idea is that if you see a lot of incorrect password attempts, you also know that something funny is up. Or, you could also see that a login attempt was issued from a computer in Nigeria (even with correct credentials). Imagine if you got a popup on your phone whenever you logged into your bank account. Wouldnt you feel a lot more safe that no one else has your information?
  - glurgh 13y ago
    The scenario presented starts with a stolen db. Also it's about recording successful, rather than unsuccessful logins. Gmail is a good counter-example - it does tell you of your last login but who keeps track of these? You log in from home, from work, from your iPhone, from your friend's wifi, etc. And you needn't stop there. What if the auditing service is compromised? What if the convenience-inclined user is using the same password at myfancyaudit.com as they are at mydogfriends.com and at their bank? Maybe I'm missing something but it mostly seems like a way of shuffling trust around without significantly improving security.
    - chacham15 13y ago
      Its not shuffling around security and neither is it adding any (at least in the crypto sense). Rather, it is a post-mortem way of knowing whether something is wrong. Assume that neither the auditing agency is cracked nor the original service. If your password gets phished and the attacker logs in as you, you will get a notification about it. Then, you can at least do something retroactively (even seconds later if you get a notification on your phone) to prevent further injury instead of finding out weeks later when all the damage is already done.
PaperclipTaken 13y ago
There is a solution that is easier than auditing, found within the public key system. You could use the same password for every single website without any of them ever knowing what it is. You would instead associate your account with a public key and then use it to verify your identity every time you wanted to log in. Then, the only vulnerability is your local machine. If someone hacks a website, the only password related information they can access is your public key, but you tell that to everyone anyway. They won't be able to use that to log onto any other website, even though you use the same password for all of them. You would still probably want to use multiple public keys, and two-factor authentication (to eliminate the single-point-of-failure risk), but the technology already exists for us to be doing this. It just needs that extra layer that will make using such a system easy for grandma, and then of course for websites to start accepting public key authentication instead of password authentication. edit: http://en.wikipedia.org/wiki/Public-key_cryptography http://en.wikipedia.org/wiki/Public-key_cryptography
- 1qaz2wsx3edc 13y ago
  It kind of saddens me that the W3C is willing to focus on things like DRM, but not a public-key implementation for browser vendors. Yeesh.
  - 0x0 13y ago
    Well, there is always the html5 <keygen> tag? http://www.w3.org/TR/html-markup/keygen.html http://www.w3.org/TR/html-markup/keygen.html https://developer.mozilla.org/en-US/docs/HTML/Element/keygen https://developer.mozilla.org/en-US/docs/HTML/Element/keygen
    - nly 13y ago
      What!? No demoscene music in my <keygen> ?
  - drdaeman 13y ago
    Actually, there's a W3C WG that works (worked?) on a protocol called WebID, that uses public keys to authenticate users, and servers are not identity providers but just serving profile information. http://www.w3.org/2005/Incubator/webid/spec/#authentication-sequence http://www.w3.org/2005/Incubator/webid/spec/#authentication-...
SeoxyS 13y ago
While I agree with the premise of the article; the author's proposal is somewhat ridiculous. It faces pretty much all the same problems as OpenID.
elchief 13y ago
Just use Google OpenID on your site, then you don't have to worry about your customers' passwords getting stolen. Simplistic? Yup. Perfect? Nope. Does the trick? Sure.
- venomsnake 13y ago
  And when google bans the account for no good reason?
shurcooL 13y ago
I long for a future where passwords are no longer necessary because the info you want to access is public anyway.
hcarvalhoalves 13y ago
> If I owned myaudithooks.com, I could have it email me based on certain rules. (...) I’d be able to sleep a little easier knowing that if one of my account credentials were compromised — whether it be due to my own carelessness or that of a random developer’s — I’d have a nice history and potentially even alerting so I know about it when it happens. ... and since it relies on DNS, it's useless.
CurtMonash 13y ago
1. This is an excellent idea, for the reasons stated. Auditing is a huge part of enterprise best practice; it should be best practice for consumers as well, and enterprises should help them with it. 2. That said, the number of sites for which we need really solid security is probably under 10. For me it's primary email, 3 for websites I own/hosting/etc., Paypal, bank, arguably Amazon, and not much else. And frankly I forget my bank password and have to keep resetting it (which is part of why my email needs to be secure).
- jacquesm 13y ago
  you're leaking a lot of valuable information here.
  - CurtMonash 13y ago
    If I can't keep my email address secure, I'm dead meat anyway. And the fact that I have a number of websites is not exactly a secret.
    - CurtMonash 13y ago
      Umm, that could more accurately have been phrased "email accounts".
pfortuny 13y ago
You might take a look here [1], using asymmetric crypto. We are still developing it but that is our approach. Yes, there is some overhead but you will need it anyway. The doc is a bit outdated but the prototypes work. [1] http://www.thesibyl.net http://www.thesibyl.net
jtheory 13y ago
There are a lot of comments touting using public keys for this. So, you have one laptop and one mobile phone. How do you set up your "password" on both? Assume you're not "technical". Next: your car (or apartment, etc.) is robbed. You lose both devices. OR: water damage. Whatever. You're away from home with one device, and its battery dies, wifi fails, whatever. No worries, your brother/friend/colleague lends you a device. But... you're locked out of EVERYTHING on that device. All of your accounts require your public key, and you obviously don't have that at an internet cafe, or even at your brother's house, on a computer you're comfortable is safe.
- claudius 13y ago
  Use the same keyring on both devices. It’s just a file, copying it over shouldn’t be a problem. Then ascii-armor that file and print it out and/or store it on a CD, keep that in a (or possibly multiple) safe location(s). You could even give your brother the key and trust him not to try to crack the passphrase.
- venomsnake 13y ago
  a few one time keys? the way google backup codes work. I am heavy RDP user. One of the things I hate about windows is the inability to have one time burn passwords.
- count 13y ago
  This is a solved problem. Look at CAC/PIV smartcards. Virtually anyone with a computer is used to carrying around a credit card or ID card. Just add a PK circuit.
  - jtheory 13y ago
    All of the problems I mentioned are solved for technical users, or special situations with the right equipment available (like a CAC reader... I certainly don't have one of those integrated into my mobile phone, or laptop for that matter). That doesn't help when we need a general solution with usability that comes close to remembering a password.
jtheory 13y ago
The first problem with myaudithooks.com is the vulnerability of the audit site, especially if this is a hosted service accepting login reports streaming from lots of different sites. I know, the actual usernames/passwords aren't part of the audit trail. Or... the passwords aren't. The usernames probably are, because otherwise how would this login report be linked to the right user? So then think about what you could do with that data, if you cracked myaudithooks.com. For the users/services using it, you'd have a comprehensive list (likely with usenames) of all of the sites these users access regularly, probably mostly with the same weak/reused passwords. Alternative: don't crack it, just ruin it. How is myaudithooks.com going to really lock down the API calls so that only mydogfriends.com can report login events to mydogfriends.com... Otherwise you could have script kiddies just throwing in lots of junk data for fun, or a malicious attempt to discredit a bank, competitor, etc. by convincing many of their users (via fake audit reports) that their accounts are cracked. I do think the general idea is good, though -- just not centralized. Login auditing is good. I always appreciate the extra bit of info when I SSH into a server and see "Last login: (timestamp) from (IP.host.blah)". It's also hardly ever used. I suspect some of the problem is that the timestamp bit would make sense to most users, but the source is more likely to be confusing for the general public. A raw IP address certainly won't be useful, and location is very scary if they don't know it's not exact and sometimes wrong ("it said last login from the next town over... where my ex lives!!!"), and that means support costs for false positives.
- jenandre 13y ago
  The usernames do not need to be part of the audit trail -- the unique URL identifies the site it came from. You know you your own logins, presumably, so it doesn't need to be there. myaudithooks.com should not be centralized -- I 100% agree. You should just be able to put any old URL there. If I want my audit entries to go to my box, I absolutely should be able to. You're right, some attacker could post bogus information if the know the url... however, the urls should be unique enough that they can't just be randomly "guessed" by an attacker. Another alternative is to provide a callback url, the way stripe does, so the service can "authenticate" the audit log entry before I consider it valid. For example, if I receive an audit log entry with id "abcd1234", I can hit http://mydogfriends.com/audit/abcd1234 http://mydogfriends.com/audit/abcd1234 to make sure it responds appropriately. The idea isn't fully fleshed out yet, but I think it wouldn't take a lot of work to make it happen.
Spearchucker 13y ago
First, it's awesome that this rather important subject is getting attention, and kudos to Jen for giving it thought and providing a perspective [1]. Logging adds value once an account is compromised, so is a contingency measure (minimising impact). Solving this problem requires mitigation (minimising probability) as well as contingency. Many comments here suggest the use of PKI. PKI is part of a solution and, as with logging and audits, requires more. PKI's biggest problem is that it doesn't scale. The question then becomes one of issuance. Self-issued certificates have no credibility so cannot be trusted. That problem is solved by using a certificate authority that can be trusted. That leads to the next question, which is whom to appoint as a certificate authority (CA)? This thing [2] should scale, so my certificate must work world-wide, with any web site and, while we're at it, offline in meat space. That makes canidates like my bank or a credit check agency bad choices, because they haven't the reach. Credit card companies like Visa, AMEX or Mastercard have better reach, but that only accounts for a tiny percentage of the world's population, because it excludes the poor. The only viable choice then becomes the organisation that already deals in identity - government departments that issue passports, ID cards and driving licences. Of course that's just national. The solution to that problem is to federate identities between governments. Awesome. First part of the problem solved [3]. The next part of the problem is anonymity, and it's various variations. This can be solved by using a privacy-enhancing credential - a problem that was solved by Stefan Brands at Credentica [4]. The tradgedy is that while these solutions exist, are mature and proven, there is just not enough incentive to make them a reality. Some closing notes: Any good identity system must adhere to all the laws of identity (http://www.identityblog.com/?p=352/#lawsofiden_topic3 http://www.identityblog.com/?p=352/#lawsofiden_topic3). Technologies like OpenId, Persona and OAuth don't even come close. SAML and WS-Federation do a much better job of it, but both SOAP/XML-based. Which makes them unpalatable to many. There's an oppotunity in there somewhere... [1] Logging has been done - just not at the scale Jens envisions. Many military and intelligence systems inform the user of previous acvitivy on login. The better ones use geolocation to tell the user where the most recent logins occured. [2] "This thing" is really an identity meta system. [3] This technology exists in the form of WS-Federation and, to a lesser extent, SAML (which relies on the browser, so doesn't work at the service level). [4] Stefan Brands came up with U-Prove, which was open-sourced by Microsoft after they aquired the technology. IBM have also come up with a privacy-enhancing technology (PET) called idemix (identity mixer). An important attribute of any self-respecting PET is that it must be claims-based, so that it functions in a world of federated identities.
EGreg 13y ago
Or there is also this solution: http://www.faqs.org/patents/app/20120110469 http://www.faqs.org/patents/app/20120110469 What do you guys think of it?
delinka 13y ago
Why are we not already using the equivalent of an RSA token for authentication on the web? Of course, the physical token is replaced with a software app on your computer or mobile device. Is there something about the physical device that cannot be replicated in software?
- drdaeman 13y ago
  Physical devices are there for a reason. If some malware gets to your phone or computer, sniffs the password and steals the keyfile it's over. There're no other options but to immediately revoke the key. Hardware security tokens are specifically meant to mitigate this issue. Otherwise, there're many software implementations out there. Your browser should already have one (search for HTTPS client certificate authentication), although it's not universal due to some X.509 PKI architecture constraints.
MarkMc 13y ago
Humans are notoriously bad at choosing passwords, so what about this approach: When you register for an account you write down on a piece of paper 4 words chosen at random by a computer (eg. "regain gauge chest Texas"). Then to log you provide (a) your email address, (b) your password, and (c) the passphrase printed on the paper. This is bit of a pain for the user, but it would greatly strengthen the security of the website because it would not depend on the security of any other websites. For a to-do-list website I can see it's not worth it, but I cannot understand how some financial websites still think it is acceptable to use only email+password authentication. (I'm looking at you, Mint.com and Schwab.com). PS: I just tried registering for a Mint.com account - it didn't let me use "password" as my password, but when I used "password1" it said, "You have a Good password". Wow.
- endianswap 13y ago
  That's just a bigger password that they're instructed to keep a copy of then? If you're going to do that, just add a proper OTP factor like a SecurID token that they can carry with them.
  - MarkMc 13y ago
    Yes a SecurID token is probably more secure than a piece of paper, but it is also less convenient and more expensive. Also, the piece of paper with 4 words will have 52 bits of entropy while a SecurID token with 6 digits has only has 20 bits of entropy. So I wonder whether SecurID would be easier for an attacker who has access to the password hash and salt to derive the original password...? In any case, whether you use a piece of paper or SecurID the main point is that financial sites which only require email+password are being negligent in their duty to protect user data from unauthorised access.
pallavkaushish 13y ago
I tried using various Single Sign On solutions but somehow their system are not satisfactory. I recently heard about this new company called SmartSignin - https://www.smartsignin.com https://www.smartsignin.com and their website says that their patented technology doesn't store your password anywhere so it's impossible to hack it. I don't know how they do it but I'm intrigued. Does anybody else know anything about them?

96 points by jenandre 13y ago

evv 13y ago

This is a clever security enhancement and wouldn't be hard for anyone to implement. Login data could also be sent over email to a specified address. We need an open standard for this.

Delphie 13y ago

Am I the only one who thinks that medium.com is abusing the HN ranking system?

rgbrenner 13y ago
in what way? they have 2 articles on the front page, 0 on the second page, at this moment.
- pi18n 13y ago
  They've had quite a few posts on the front page in the last week or so.
drivebyacct2 13y ago
That's like saying wordpress.com or blogspot.com are abusing the ranking system...
- Sprint 13y ago
  Those are not curated while medium.com is "creating content is limited to a small list [of users]".
jacquesm 13y ago
Yes, definitely. I noted as much yesterday in another comment.

eksith 13y ago

Unfortunately, passwords are to account security what PHP is to web development. It may not be the best, it's not the fastest, it even may not be the most secure. But it's ubiquitous, it's free and if done right, it's an OK way to get the job done precluding more complex solutions that aren't as practical for most of us. I think I posted this elsewhere or maybe here, but I got around the password remembering problem by keeping one master file of all passwords/usernames/emails etc... that I keep GPG encrypted ( specifically http://www.gpg4win.org/about.html http://www.gpg4win.org/about.html ). All passwords in the master file are randomly generated. That is, of course, the account I'm creating is a throwaway at Paul's Peanut Brittle shop. Interestingly, paulspeanutbrittle.com isn't taken.

hooande 13y ago

This sounds like a good idea. I think it gets one thing wrong: Hacking living social was not about gaining access to anyone's living social account. It was about getting passwords that were reused on sites that actually matter (banking, etc). A notification that someone has logged into your livingsocial|facebook|twitter|linkedin is only valuable if it compels you to immediately change all of your other passwords. Which might be a decent idea for a version 2 of this hypothetical software...automatically change all of my passwords whenever one of my accounts is compromised. The sad thing is that people reuse their passwords so often. Just remembering two passwords, one for important stuff and one for social stuff, would go a long way toward making this a non-problem and taking a security burden off of a lot of startups.

glurgh 13y ago

Wouldn't an attacker who's gained enough access to a system to get db dumps or modify serverside code to the point where they can capture incoming passwords simply turn off the audit calls? Wouldn't an attacker who's gained access to a secondary account through a password derived from a db dump change the audit URL? Beside that, knowing an account is compromised is better than nothing but often not particularly useful - see all the horror stories by people who've had their gmail account compromised. They tend to find out very quickly the account has been taken over, recovery is still difficult and the loss of data/time/neurons is often significant.

chacham15 13y ago
A user isnt only vulnerable when the db gets hacked. People have other ways at getting at passwords (sometimes they guess). The idea is that if you see a lot of incorrect password attempts, you also know that something funny is up. Or, you could also see that a login attempt was issued from a computer in Nigeria (even with correct credentials). Imagine if you got a popup on your phone whenever you logged into your bank account. Wouldnt you feel a lot more safe that no one else has your information?
- glurgh 13y ago
  The scenario presented starts with a stolen db. Also it's about recording successful, rather than unsuccessful logins. Gmail is a good counter-example - it does tell you of your last login but who keeps track of these? You log in from home, from work, from your iPhone, from your friend's wifi, etc. And you needn't stop there. What if the auditing service is compromised? What if the convenience-inclined user is using the same password at myfancyaudit.com as they are at mydogfriends.com and at their bank? Maybe I'm missing something but it mostly seems like a way of shuffling trust around without significantly improving security.
  - chacham15 13y ago
    Its not shuffling around security and neither is it adding any (at least in the crypto sense). Rather, it is a post-mortem way of knowing whether something is wrong. Assume that neither the auditing agency is cracked nor the original service. If your password gets phished and the attacker logs in as you, you will get a notification about it. Then, you can at least do something retroactively (even seconds later if you get a notification on your phone) to prevent further injury instead of finding out weeks later when all the damage is already done.

PaperclipTaken 13y ago

There is a solution that is easier than auditing, found within the public key system. You could use the same password for every single website without any of them ever knowing what it is. You would instead associate your account with a public key and then use it to verify your identity every time you wanted to log in. Then, the only vulnerability is your local machine. If someone hacks a website, the only password related information they can access is your public key, but you tell that to everyone anyway. They won't be able to use that to log onto any other website, even though you use the same password for all of them. You would still probably want to use multiple public keys, and two-factor authentication (to eliminate the single-point-of-failure risk), but the technology already exists for us to be doing this. It just needs that extra layer that will make using such a system easy for grandma, and then of course for websites to start accepting public key authentication instead of password authentication. edit: http://en.wikipedia.org/wiki/Public-key_cryptography http://en.wikipedia.org/wiki/Public-key_cryptography

1qaz2wsx3edc 13y ago
It kind of saddens me that the W3C is willing to focus on things like DRM, but not a public-key implementation for browser vendors. Yeesh.
- 0x0 13y ago
  Well, there is always the html5 <keygen> tag? http://www.w3.org/TR/html-markup/keygen.html http://www.w3.org/TR/html-markup/keygen.html https://developer.mozilla.org/en-US/docs/HTML/Element/keygen https://developer.mozilla.org/en-US/docs/HTML/Element/keygen
  - nly 13y ago
    What!? No demoscene music in my <keygen> ?
- drdaeman 13y ago
  Actually, there's a W3C WG that works (worked?) on a protocol called WebID, that uses public keys to authenticate users, and servers are not identity providers but just serving profile information. http://www.w3.org/2005/Incubator/webid/spec/#authentication-sequence http://www.w3.org/2005/Incubator/webid/spec/#authentication-...

SeoxyS 13y ago

While I agree with the premise of the article; the author's proposal is somewhat ridiculous. It faces pretty much all the same problems as OpenID.

elchief 13y ago

Just use Google OpenID on your site, then you don't have to worry about your customers' passwords getting stolen. Simplistic? Yup. Perfect? Nope. Does the trick? Sure.

venomsnake 13y ago
And when google bans the account for no good reason?

shurcooL 13y ago

I long for a future where passwords are no longer necessary because the info you want to access is public anyway.

hcarvalhoalves 13y ago

> If I owned myaudithooks.com, I could have it email me based on certain rules. (...) I’d be able to sleep a little easier knowing that if one of my account credentials were compromised — whether it be due to my own carelessness or that of a random developer’s — I’d have a nice history and potentially even alerting so I know about it when it happens. ... and since it relies on DNS, it's useless.

CurtMonash 13y ago

1. This is an excellent idea, for the reasons stated. Auditing is a huge part of enterprise best practice; it should be best practice for consumers as well, and enterprises should help them with it. 2. That said, the number of sites for which we need really solid security is probably under 10. For me it's primary email, 3 for websites I own/hosting/etc., Paypal, bank, arguably Amazon, and not much else. And frankly I forget my bank password and have to keep resetting it (which is part of why my email needs to be secure).

jacquesm 13y ago
you're leaking a lot of valuable information here.
- CurtMonash 13y ago
  If I can't keep my email address secure, I'm dead meat anyway. And the fact that I have a number of websites is not exactly a secret.
  - CurtMonash 13y ago
    Umm, that could more accurately have been phrased "email accounts".

pfortuny 13y ago

You might take a look here [1], using asymmetric crypto. We are still developing it but that is our approach. Yes, there is some overhead but you will need it anyway. The doc is a bit outdated but the prototypes work. [1] http://www.thesibyl.net http://www.thesibyl.net

jtheory 13y ago

There are a lot of comments touting using public keys for this. So, you have one laptop and one mobile phone. How do you set up your "password" on both? Assume you're not "technical". Next: your car (or apartment, etc.) is robbed. You lose both devices. OR: water damage. Whatever. You're away from home with one device, and its battery dies, wifi fails, whatever. No worries, your brother/friend/colleague lends you a device. But... you're locked out of EVERYTHING on that device. All of your accounts require your public key, and you obviously don't have that at an internet cafe, or even at your brother's house, on a computer you're comfortable is safe.

claudius 13y ago
Use the same keyring on both devices. It’s just a file, copying it over shouldn’t be a problem. Then ascii-armor that file and print it out and/or store it on a CD, keep that in a (or possibly multiple) safe location(s). You could even give your brother the key and trust him not to try to crack the passphrase.
venomsnake 13y ago
a few one time keys? the way google backup codes work. I am heavy RDP user. One of the things I hate about windows is the inability to have one time burn passwords.
count 13y ago
This is a solved problem. Look at CAC/PIV smartcards. Virtually anyone with a computer is used to carrying around a credit card or ID card. Just add a PK circuit.
- jtheory 13y ago
  All of the problems I mentioned are solved for technical users, or special situations with the right equipment available (like a CAC reader... I certainly don't have one of those integrated into my mobile phone, or laptop for that matter). That doesn't help when we need a general solution with usability that comes close to remembering a password.

jtheory 13y ago

The first problem with myaudithooks.com is the vulnerability of the audit site, especially if this is a hosted service accepting login reports streaming from lots of different sites. I know, the actual usernames/passwords aren't part of the audit trail. Or... the passwords aren't. The usernames probably are, because otherwise how would this login report be linked to the right user? So then think about what you could do with that data, if you cracked myaudithooks.com. For the users/services using it, you'd have a comprehensive list (likely with usenames) of all of the sites these users access regularly, probably mostly with the same weak/reused passwords. Alternative: don't crack it, just ruin it. How is myaudithooks.com going to really lock down the API calls so that only mydogfriends.com can report login events to mydogfriends.com... Otherwise you could have script kiddies just throwing in lots of junk data for fun, or a malicious attempt to discredit a bank, competitor, etc. by convincing many of their users (via fake audit reports) that their accounts are cracked. I do think the general idea is good, though -- just not centralized. Login auditing is good. I always appreciate the extra bit of info when I SSH into a server and see "Last login: (timestamp) from (IP.host.blah)". It's also hardly ever used. I suspect some of the problem is that the timestamp bit would make sense to most users, but the source is more likely to be confusing for the general public. A raw IP address certainly won't be useful, and location is very scary if they don't know it's not exact and sometimes wrong ("it said last login from the next town over... where my ex lives!!!"), and that means support costs for false positives.

jenandre 13y ago
The usernames do not need to be part of the audit trail -- the unique URL identifies the site it came from. You know you your own logins, presumably, so it doesn't need to be there. myaudithooks.com should not be centralized -- I 100% agree. You should just be able to put any old URL there. If I want my audit entries to go to my box, I absolutely should be able to. You're right, some attacker could post bogus information if the know the url... however, the urls should be unique enough that they can't just be randomly "guessed" by an attacker. Another alternative is to provide a callback url, the way stripe does, so the service can "authenticate" the audit log entry before I consider it valid. For example, if I receive an audit log entry with id "abcd1234", I can hit http://mydogfriends.com/audit/abcd1234 http://mydogfriends.com/audit/abcd1234 to make sure it responds appropriately. The idea isn't fully fleshed out yet, but I think it wouldn't take a lot of work to make it happen.

Spearchucker 13y ago

First, it's awesome that this rather important subject is getting attention, and kudos to Jen for giving it thought and providing a perspective [1]. Logging adds value once an account is compromised, so is a contingency measure (minimising impact). Solving this problem requires mitigation (minimising probability) as well as contingency. Many comments here suggest the use of PKI. PKI is part of a solution and, as with logging and audits, requires more. PKI's biggest problem is that it doesn't scale. The question then becomes one of issuance. Self-issued certificates have no credibility so cannot be trusted. That problem is solved by using a certificate authority that can be trusted. That leads to the next question, which is whom to appoint as a certificate authority (CA)? This thing [2] should scale, so my certificate must work world-wide, with any web site and, while we're at it, offline in meat space. That makes canidates like my bank or a credit check agency bad choices, because they haven't the reach. Credit card companies like Visa, AMEX or Mastercard have better reach, but that only accounts for a tiny percentage of the world's population, because it excludes the poor. The only viable choice then becomes the organisation that already deals in identity - government departments that issue passports, ID cards and driving licences. Of course that's just national. The solution to that problem is to federate identities between governments. Awesome. First part of the problem solved [3]. The next part of the problem is anonymity, and it's various variations. This can be solved by using a privacy-enhancing credential - a problem that was solved by Stefan Brands at Credentica [4]. The tradgedy is that while these solutions exist, are mature and proven, there is just not enough incentive to make them a reality. Some closing notes: Any good identity system must adhere to all the laws of identity (http://www.identityblog.com/?p=352/#lawsofiden_topic3 http://www.identityblog.com/?p=352/#lawsofiden_topic3). Technologies like OpenId, Persona and OAuth don't even come close. SAML and WS-Federation do a much better job of it, but both SOAP/XML-based. Which makes them unpalatable to many. There's an oppotunity in there somewhere... [1] Logging has been done - just not at the scale Jens envisions. Many military and intelligence systems inform the user of previous acvitivy on login. The better ones use geolocation to tell the user where the most recent logins occured. [2] "This thing" is really an identity meta system. [3] This technology exists in the form of WS-Federation and, to a lesser extent, SAML (which relies on the browser, so doesn't work at the service level). [4] Stefan Brands came up with U-Prove, which was open-sourced by Microsoft after they aquired the technology. IBM have also come up with a privacy-enhancing technology (PET) called idemix (identity mixer). An important attribute of any self-respecting PET is that it must be claims-based, so that it functions in a world of federated identities.

EGreg 13y ago

Or there is also this solution: http://www.faqs.org/patents/app/20120110469 http://www.faqs.org/patents/app/20120110469 What do you guys think of it?

delinka 13y ago

Why are we not already using the equivalent of an RSA token for authentication on the web? Of course, the physical token is replaced with a software app on your computer or mobile device. Is there something about the physical device that cannot be replicated in software?

drdaeman 13y ago
Physical devices are there for a reason. If some malware gets to your phone or computer, sniffs the password and steals the keyfile it's over. There're no other options but to immediately revoke the key. Hardware security tokens are specifically meant to mitigate this issue. Otherwise, there're many software implementations out there. Your browser should already have one (search for HTTPS client certificate authentication), although it's not universal due to some X.509 PKI architecture constraints.

MarkMc 13y ago

Humans are notoriously bad at choosing passwords, so what about this approach: When you register for an account you write down on a piece of paper 4 words chosen at random by a computer (eg. "regain gauge chest Texas"). Then to log you provide (a) your email address, (b) your password, and (c) the passphrase printed on the paper. This is bit of a pain for the user, but it would greatly strengthen the security of the website because it would not depend on the security of any other websites. For a to-do-list website I can see it's not worth it, but I cannot understand how some financial websites still think it is acceptable to use only email+password authentication. (I'm looking at you, Mint.com and Schwab.com). PS: I just tried registering for a Mint.com account - it didn't let me use "password" as my password, but when I used "password1" it said, "You have a Good password". Wow.

endianswap 13y ago
That's just a bigger password that they're instructed to keep a copy of then? If you're going to do that, just add a proper OTP factor like a SecurID token that they can carry with them.
- MarkMc 13y ago
  Yes a SecurID token is probably more secure than a piece of paper, but it is also less convenient and more expensive. Also, the piece of paper with 4 words will have 52 bits of entropy while a SecurID token with 6 digits has only has 20 bits of entropy. So I wonder whether SecurID would be easier for an attacker who has access to the password hash and salt to derive the original password...? In any case, whether you use a piece of paper or SecurID the main point is that financial sites which only require email+password are being negligent in their duty to protect user data from unauthorised access.

pallavkaushish 13y ago

I tried using various Single Sign On solutions but somehow their system are not satisfactory. I recently heard about this new company called SmartSignin - https://www.smartsignin.com https://www.smartsignin.com and their website says that their patented technology doesn't store your password anywhere so it's impossible to hack it. I don't know how they do it but I'm intrigued. Does anybody else know anything about them?