Slashdot Mirror


Why Are We Still Using Passwords? (securityledger.com)

Here's some surprising news from the Akamia Edge conference. chicksdaddy writes: [E]xecutives at some of the U.S.'s leading corporations agreed that the much maligned password won't be abandoned any time soon, even as data breaches and follow-on attacks make passwords more susceptible than ever to abuse, the Security Ledger reports. "We reached the end of needing passwords maybe seven years ago, but we still use them," said Steve Winterfeld, Director of Cybersecurity, at clothing retailer Nordstrom. "They're still the primary layer of defense."

"It's hard to kill them," noted Shalini Mayor, who is a Senior Director at Visa Inc. "The question is what to replace them with." This, even though the cost of using passwords is high and getting higher, as sophisticated attacks attempt to compromise legitimate accounts using so-called "credential stuffing" techniques, which use automated password guessing attacks against web-based applications... Stronger and more reliable alternatives to passwords already exist, but the obstacles to using them are often prohibitive. Shalani Mayor said Visa is "looking at" biometric technologies like Apple's TouchID as a tool for making payments securely. Such technologies -- from fingerprint scans to facial and retinal scans -- promise more secure and reliable factors than alphanumeric passwords, the executives agreed. But customers often resist the technologies or find them error prone or too difficult to use.

139 of 209 comments (clear)

  1. Those... arenâ(TM)t more secure by Anonymous Coward · · Score: 4, Insightful

    Biometrics are not more secure than passwords - theyâ(TM)re less secure but sufficiently more convenient that you can convince people to use them.

    We still use passwords because theyâ(TM)re still the most secure way of authenticating your identity when combined with a second factor.

    1. Re:Those... arenâ(TM)t more secure by nitehawk214 · · Score: 5, Funny

      Like passwords with unicode in them. Impossible to share via Slashdot.

      --
      I'm a good cook. I'm a fantastic eater. - Steven Brust
    2. Re:Those... arenâ(TM)t more secure by fahrbot-bot · · Score: 5, Informative

      Biometrics are not more secure than passwords - they're less secure but sufficiently more convenient that you can convince people to use them.

      A fingerprint is more convenient until the moment you get a blister (or some other damage) on your finger(s), then you're locked out. Seems unlikely? When I got a job at the NASA LaRC way, way back, I had to get fingerprinted, but couldn't because I had been working on my car that week and my hands and fingers were all beat up. I had to wait a week for them to clear up enough to get processed.

      --
      It must have been something you assimilated. . . .
    3. Re:Those... arenâ(TM)t more secure by rastos1 · · Score: 2

      Biometrics are not more secure than passwords ...

      I thought biometrics also serves a different purpose - identification, rather than authentication.

    4. Re:Those... arenâ(TM)t more secure by ShanghaiBill · · Score: 1

      A fingerprint is more convenient until the moment you get a blister (or some other damage) on your finger(s), then you're locked out.

      Every biometric system I use has a password fallback. The biometric is only for convenience.

    5. Re:Those... arenâ(TM)t more secure by ShanghaiBill · · Score: 1

      I thought biometrics also serves a different purpose - identification, rather than authentication.

      My laptop uses it for both. My fingerprint authenticates me and gives me access. It also identifies me, so it opens the right account. If my wife puts her fingerprint on the scanner instead, it opens her account.

      My bank uses it only for identification. I still need to show an ID or enter a PIN for most transactions.

    6. Re:Those... arenâ(TM)t more secure by Wycliffe · · Score: 1

      We still use passwords because theyâ(TM)re still the most secure way of authenticating your identity when combined with a second factor.

      About a decade ago, Etrade sent me a small free keychain about the size of a stick of gum (1cm x 4cm x 1/2cm). It had a small digital display that had a password that changed every 60 seconds and was somehow synced with etrade's webserver. Even without a secondary password, this is a very secure system. It's not connected to the internet and you would likely know immediately if it was stolen. It ran on a single watch battery for over a year. It likely only cost a couple dollars to produce. The technology exists and has existed for a long time. The problem is that for whatever reason, people would rather use the same password for 20 websites than to have to look at their keychain and type in a 6 digit number every time they want to log in.

    7. Re: Those... arenâ(TM)t more secure by Monster_user · · Score: 1

      That keychain is usually a part of a 2-factor authentication system for a website continuing high value or high risk information.

      Many sites have apps for a smart phone which do the same thing. Otherwise you need a keychain for every site.

    8. Re: Those... arenâ(TM)t more secure by ncc74656 · · Score: 1

      Many sites have apps for a smart phone which do the same thing. Otherwise you need a keychain for every site.

      ...or a device that supports a standard like U2F. There are at least a couple of options here; the Trezor that holds my Bitcoin stash is also currently set up as a second factor for Google and Dropbox. There's also a YubiKey that supports U2F.

      --
      20 January 2017: the End of an Error.
    9. Re:Those... arenâ(TM)t more secure by redmid17 · · Score: 1

      No one tell this guy about RSA

    10. Re:Those... arenâ(TM)t more secure by Wycliffe · · Score: 1

      No one tell this guy about RSA

      Yes, it was manufactured by RSA. That's not really the point. The point is that it has existed for a decade, it doesn't require an internet connection, and they are cheap to produce. There are even free versions today that use an app on your phone. Noone wants to use them. Most people would prefer to either use the same password everywhere or to have all their passwords memorized so if someone steals your phone they automatically have all your passwords too. People don't care about security until after something bad happens.

    11. Re:Those... arenâ(TM)t more secure by lsatenstein · · Score: 1

      Perhaps I have to use a touch pad that punctures my finger and takes a blood sample. It will check the dna of my sample, and if it matches what is on file, I will be able to log into the system. Perhaps, in lieu of the blood sample, a saliva sample would suffice. Blow into a straw and if the dna matches, voila-- access.

      --
      Leslie Satenstein Montreal Quebec Canada
    12. Re:Those... arenâ(TM)t more secure by stoatwblr · · Score: 1

      "When I got a job at the NASA LaRC way, way back, I had to get fingerprinted, but couldn't because I had been working on my car that week and my hands and fingers were all beat up. I had to wait a week for them to clear up enough to get processed."

      Or you could be in the unfortunate position my wife constantly finds herself in when faced with biometric demands for fingerprints (primarily immigration/visa issues) - her fingerprints are so light that most systems simply can't pick them up. (She can't use iphone/android fingerprint sensors either)

    13. Re:Those... arenâ(TM)t more secure by KingBenny · · Score: 1

      yah, i dont get this either, it must be something like 3D-tv and smart watches, someone sold the pitch to a suit who doesnt get shit but money
      biometrics would be static data just like a fucking password if you'll pardon my american english im having a headache for a week now
      and that means all you have to is get to that, so if you can decrypt the password then why the fuck would you not be able to decrypt the static data, the "unique" id stored as ones and zeros ... or even ones, zeroes and onezerous in a qubit (someone still has to explain exactly to me why a qubit is different from a terniary, all i ever hear is it has three states so is the third state of one&&zero at once different from a terniary third state ? if so, how ... wooooooffftopic sorry)
      thats how secure it is, right ? its static data stored encrypted
      like a fucking password ?
      or am i talking out of my crank uneducated, o no, un-DEGREEd ass again ?
      its the marketeers thats the problem and the suits who dont get shit but money, thats where the vulnerability is

      --
      Free speech was meant to be free for all... how can anyone grow up in a nanny state ?
    14. Re:Those... arenâ(TM)t more secure by KingBenny · · Score: 1

      here i go again answering myself but you can blame the headache and down the line whoever gave that to me ... i think OTP (like yubikey in essence) or like a digipass i get here from my bank would be lot better to start but still there will be weak points i suppose since all you need is the "all clear" signal in any case of unlocking anything if you look at it from a metaverse perspective

      --
      Free speech was meant to be free for all... how can anyone grow up in a nanny state ?
  2. Biometrics arenâ(TM)t secure... by Anonymous Coward · · Score: 5, Insightful

    If I ever get arrested or stopped at the airport, my phone could be unlocked by forcing my finger on the button or scanning my face(iPhone X). So without a password, biometrics can trivially compromise your security against state/pseudostate actors when they have physical access to you. At least with a passcode they have to observe some sort of due process to coerce you.

    1. Re:Biometrics arenâ(TM)t secure... by Type44Q · · Score: 2

      I've never heard of a pipe wrench referred to as "due process" before.

    2. Re:Biometrics arenâ(TM)t secure... by myowntrueself · · Score: 1

      I've never heard of a pipe wrench referred to as "due process" before.

      Quite correct. For it to count as "due process" it has to not leave a mark. Something like waterboarding for example.

      --
      In the free world the media isn't government run; the government is media run.
    3. Re:Biometrics arenâ(TM)t secure... by redmid17 · · Score: 1

      So do not use your finger or face to log into your phone?

      Fucking mindblowing I know

  3. customers often resist the technologies by Jamlad · · Score: 4, Insightful
    because as everyone with half a brain realizes that biometrics are a fucktarded method of authentication. A keyword gets exposed, fine. Change it. Your fingerprint gets exposed? How are you going to revise that?

    The best method of authentication, as far I I've experienced, is a physical token (keycard). Worst case scenario, I don't notice it's missing after two days (Friday evening till Monday morning). Chances are I've dropped in a city centre rather than haven it exploited by an unknown agency. Even still, they;ve only got the physical credentials of a low-tier employee. On-site physical access is still required.

    1. Re:customers often resist the technologies by Greyfox · · Score: 1

      Yeah, I worked at a company where you just stuck your card into whatever computer you sat down at and it would find your session out on the network and bring it to that computer. You still used a password to unlock the session, though. Without the card, your password was useless. Without your password, the card was useless. They also didn't have the fucktarded password requirements that most companies do, so you could use a passphrase, which can be significantly easier to remember and more secure than the usual corporate password.

      --

      I'm trying to teach myself to set people on fire with my mind... Is it hot in here?

    2. Re:customers often resist the technologies by djinn6 · · Score: 1

      What about a physical token with password or pin entered on the token itself, which then signs a message using its private key? The attacker would need to both observe you using that token and obtain the token itself.

    3. Re:customers often resist the technologies by houghi · · Score: 1

      The thing with a keycard is that I would need to walk around with several hundreds of them. One for each place I want access. Using a password manager would mean that if that is hacked, all the rest is hacked.

      --
      Don't fight for your country, if your country does not fight for you.
  4. "Why Are We Still Using Passwords?" by Anonymous Coward · · Score: 2, Insightful

    "We reached the end of needing passwords maybe seven years ago" - "The question is what to replace them with."

    qed

    1. Re:"Why Are We Still Using Passwords?" by Hardness · · Score: 2
    2. Re: "Why Are We Still Using Passwords?" by Zero__Kelvin · · Score: 1

      That is supposed to be a joke, right? Tell me that Gibson isn't seriously trying to say that a system where anyone can log in if they have your device, and nobody can log in from a different device, is something anyone should even consider.

      --
      Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
    3. Re: "Why Are We Still Using Passwords?" by tepples · · Score: 1

      and nobody can log in from a different device

      Easy. Associate each of your devices to your identity provider as you buy them.

    4. Re: "Why Are We Still Using Passwords?" by nitehawk214 · · Score: 1

      They key would have to be stored somewhere off of your device, otherwise you would be locked out of your own accounts if the device were to be lost or broken. This is exactly as secure as just having a private key file somewhere. With all the pitfalls involved with that.

      I see it as a step backwards from something like LastPass as it provides no security benefits, with a bunch of downfalls that you stated.

      Gibson is a smart guy, but I have no idea what problem he thought he was solving here.

      --
      I'm a good cook. I'm a fantastic eater. - Steven Brust
    5. Re: "Why Are We Still Using Passwords?" by Dutch+Gun · · Score: 1

      Even though I use it myself, I consider LastPass to be a band-aid for a system that's broken by design.

      SQRL is an attempt to solve the lunacy of requiring complex, unique passwords required for every different website you visit. We've seen time and time again that people are terrible about managing passwords. SQRL uses a secure authentication mechanism that improves on username/password in several key ways. The most important feature, IMO, is that the website only requires the use of public information to authenticate you (a public key + website URL). This means that your credentials can't be stolen by a breach of the website.

      As to the backup key, yes, you have to carefully protect it (designed as a QR code that can be printed out and stored offline), but who doesn't have access to a safety deposit box or document safe? And the code only needs to be used in case of a compromise of some sort. These days, storing a key electronically, especially on an internet-connected computer, is *vastly* more risky than even storing a piece of paper in your desk. The world's threat models have changed.

      --
      Irony: Agile development has too much intertia to be abandoned now.
    6. Re: "Why Are We Still Using Passwords?" by Zero__Kelvin · · Score: 1

      And again ... locked in to access based on using a phone. Did it ever occur to you that some people may want to use a friends computer or other device?

      --
      Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
    7. Re: "Why Are We Still Using Passwords?" by Zero__Kelvin · · Score: 1

      Again, it is a woefully failed attempt.

      --
      Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
    8. Re: "Why Are We Still Using Passwords?" by Dutch+Gun · · Score: 1

      Maybe so, but given that we haven't even seen SQRL out of the gate with an official client release, I think it's a big too early to tell. There's a lot of inertia behind password-based logins, so I'd say the odds are long against it, even if it has some good technical merits.

      FWIW, you made two completely factually-incorrect statements about SQRL in your first post. You seem a bit dismissive of the project in general, so no offense, but I'm not going to waste my time explaining further.

      --
      Irony: Agile development has too much intertia to be abandoned now.
    9. Re: "Why Are We Still Using Passwords?" by Zero__Kelvin · · Score: 1

      I didn't add "... without introducing complexity people can't understand and won't use" in my OP, as that should be assumed by any moderately competent security professional to kill the idea. Given that caveat feel free to make your case. I'd love to hear what I missed actually.

      --
      Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
    10. Re: "Why Are We Still Using Passwords?" by Dutch+Gun · · Score: 1

      without introducing complexity people can't understand and won't use

      Totally agree, but given that you pretty much need to use a password manager to safely use passwords online these days, I'm not sure how much worse a SQRL client could be. I can't vouch for ease of use, though, as I haven't actually tried the client myself yet - only heard it described in some detail.

      where anyone can log in if they have your device

      This is incorrect. The client is designed to always require a subset of your password, like a PIN, every time you authorize. More to the point, I certainly hope by now that most people have locking phones with reasonably good security.

      and nobody can log in from a different device.

      Also incorrect. Even if a thief steals your device and somehow gains access (let's assume you used a client that didn't log in, and didn't lock your device for some bizarre reason), you still have a master override code that the client prints out for you when setting up your device, with the idea being to store it in a save place (document safe, safety deposit box, etc). Using this code, you can lock out the old authentication by creating a new secret authentication key. The old authentication will be deprecated, as part of the protocol specs. Note that this isn't possible to do just with the original stolen credentials.

      --
      Irony: Agile development has too much intertia to be abandoned now.
    11. Re: "Why Are We Still Using Passwords?" by tepples · · Score: 1

      As described on the SQRL page, the procedure is as follows:

      1. Install and configure the SQRL authenticator app on your smartphone or tablet.
      2. Carry your smartphone or tablet near your friend's computer.
      3. Connect your smartphone or tablet to the Internet.
      4. Navigate to the SQRL login form on your friend's computer.
      5. Open the SQRL authenticator app on your smartphone or tablet.
      6. Inside the app, photograph the display of your friend's computer using the camera on your smartphone or tablet.
      7. In seconds, your session is authenticated, and the login form disappears.

    12. Re: "Why Are We Still Using Passwords?" by Zero__Kelvin · · Score: 1

      Someone should invent Google Authenticator.

      --
      Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
    13. Re: "Why Are We Still Using Passwords?" by tepples · · Score: 1

      SQRL and TOTP, the protocol used by Google Authenticator, both have their advantages.

      SQRL works without rekeying TOTP requires the user to look at the device, read a code, and retype it. SQRL allows the authentication request to be transmitted from the display of the friend's computer to the configured authenticator without needing to retype anything. TOTP works with an offline authenticator SQRL requires the authenticator app to have access to the Internet. The TOTP authenticator does not need to connect to the Internet during authentication; the device only needs to synchronize its time-of-day clock every so often. This is helpful for a Wi-Fi-only tablet or for a smartphone with limited or no cellular Internet data transfer allowance.
    14. Re: "Why Are We Still Using Passwords?" by Zero__Kelvin · · Score: 1

      There is no *real* advantage (OMFG! I had ta typez da numba!) and a very real disadvantage (Poor cell service or unpaid cell bill means you are locked out) in other words.

      --
      Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
    15. Re: "Why Are We Still Using Passwords?" by tepples · · Score: 1

      Poor cell service or unpaid cell bill means you are locked out

      Join Wi-Fi and you're no longer locked out.

    16. Re: "Why Are We Still Using Passwords?" by Zero__Kelvin · · Score: 1

      Not everyone has WiFi I'm afraid. Still locked out.

      --
      Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
    17. Re: "Why Are We Still Using Passwords?" by tepples · · Score: 1

      3. Connect your smartphone or tablet to the Internet.

      Not everyone has WiFi I'm afraid. Still locked out.

      If cellular is unavailable, and WI-Fi is unavailable, the remaining option I can think of is wired networking. This can be done in one of three ways:

      A. Plug a USB OTG NIC into the phone and an open Ethernet jack.
      B. Plug a battery-powered access point into an open Ethernet jack.
      C. Plug a USB cable into the phone and friend's computer, tell the phone to emulate a NIC, and tell the friend's computer to bridge the connections. Here, the phone is using the PC's Internet connection, unlike tethering which is the other way around.

      As I wrote earlier, I concede an advantage of TOTP over SQRL in situations where neither the cellular network nor the existing WLAN nor a battery-powered WLAN nor a USB NIC nor sharing of the PC's connection is possible. But how often will end users encounter all five of these situations?

    18. Re: "Why Are We Still Using Passwords?" by nitehawk214 · · Score: 1

      Even if this is implemented perfectly, I think this will go just as well as OpenId. Eventually the providers give up on it or go out of business because there is no money to be made on it, and things get fucked up. I am not going to trust my credentials to a 3rd party like that again.

      --
      I'm a good cook. I'm a fantastic eater. - Steven Brust
  5. Who and what do you trust? by hey! · · Score: 1

    You are right of course to distrust your own mind; it has a bias for convenience. But someone gives you a thing like a crypto token and tells you to entrust your deepest secrets, perhaps even to imbue the artifact with your personal authority.

    Should you trust that thing so much, keeping mind that in effect means trusting everyone involved in its programming and provisioning?

    I foresee passwords remaining useful and indeed essential, despite their obvious limitations, as part of two factor authentication. Even if you are using biometrics, those can be stolen or counterfeited by various methods.

    --
    Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
    1. Re: Who and what do you trust? by hey! · · Score: 1

      Sanders lost to Clinton because Clinton had spent nearly 40 years, from her days as First Lady of Arkansas, cultivating relationships with Democratic activists and party insiders. She'd campaigned nationally multiple times, twice as the wife of Bill Clinton and once as a candidate herself. The only way to overcome that kind of advantage is to have a candidate with immense natural political talent, like Obama.

      To win the nomination of either party you have to win the party base, which Sanders failed to do; whether you voted for Sander or Clinton was largely a function of how long you'd spent in the Democratic party. To win the generals you need someone with the power to talk to people who don't pay attention; what you want is someone who has the capacity to "come out of nowhere" and win, because that's where you're starting with the general electorate.

      People have difficulty distinguishing between their personal feelings for or against a politician and that politician's ability to win over people who aren't committed one way or the other. I supported Sanders in the primary, even though I thought Clinton would have been a better president, because I didn't think she could win.

      --
      Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
  6. Give up anonymity if it saves just one life by elainerd · · Score: 5, Insightful

    Clearly we need to replace passwords with a chip or mark or tattoo in the palms of the hands and on the foreheads / retinas, etc. Then we need to make sure that people can't buy or sell without taking these marks on themselves. Naturally cash will have to be eliminated. This way we can control and identify what the people spend their money on and we can use this information to further oppress and bind them down into abject bondage and suffering. Yep, that's the ticket. No more anonymity, all must bow down and accept the will of Evil. Every citizen a slave.

      "A jackboot stamping on a human face forever"-Orwell or Huxley, i forget and am too lazy to search.

    --
    Faith: Belief in Truth. Superstition: Belief in Falsehood.
    1. Re:Give up anonymity if it saves just one life by nitehawk214 · · Score: 1

      We live in both Huxley and Orwell's worlds at the same time.

      http://www.zerohedge.com/sites...

      --
      I'm a good cook. I'm a fantastic eater. - Steven Brust
    2. Re:Give up anonymity if it saves just one life by Ol+Olsoc · · Score: 3, Insightful

      Clearly we need to replace passwords with a chip or mark or tattoo in the palms of the hands and on the foreheads / retinas, etc. Then we need to make sure that people can't buy or sell without taking these marks on themselves. Naturally cash will have to be eliminated. This way we can control and identify what the people spend their money on and we can use this information to further oppress and bind them down into abject bondage and suffering. Yep, that's the ticket. No more anonymity, all must bow down and accept the will of Evil. Every citizen a slave.

      I had no idea that Pat Robertson was on Slashdot! You forgot the part where God will fix this problem if we just get rid of all the "homos".

      --
      The shepherds did so well protecting the flock that the sheep no longer believed that wolves existed.
    3. Re:Give up anonymity if it saves just one life by sysrammer · · Score: 1

      Good post. I have a small critique, though.

      You forgot to think of the children.

      --
      His ignorance covered the whole earth like a blanket, and there was hardly a hole in it anywhere. - Mark Twain
    4. Re:Give up anonymity if it saves just one life by WallyL · · Score: 1

      Whoosh!

  7. Oh FFS, not this crap again by Anonymous Coward · · Score: 5, Insightful

    We use passwords because it's something you know AND SOMETHING YOU CAN CHANGE WHEN COMPROMISED.

    You cannot change your fingerprints or other biometric data so when it's compromised or when technology advances in a way which allows the biometric sensors to be fooled then you are completely and totally stuffed. :-(

    Do the people proposing this ever have _ANY_ real world experience at all ?

    Oh, and yes, using biometric data allows intelligence agencies, who will likely be able to obtain that information in various ways, to pretend to be you when they want to compromise systems you control.

    1. Re:Oh FFS, not this crap again by swillden · · Score: 1

      You cannot change your fingerprints or other biometric data so when it's compromised

      This is irrelevant. I wrote a detailed explanation here: http://divegeekstuff.blogspot....

      Do the people proposing this ever have _ANY_ real world experience at all?

      I do, about 30 years' worth, in both physical and information security.

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
    2. Re:Oh FFS, not this crap again by bingoUV · · Score: 1

      Ok, read your blog post*, does not give any evidence for it being irrelevant. Your blog post says security of fingerprint authentication depends on the value of the thing being secured. Here, the AC's post as well as TFA does not specify any particular value of the value of the thing being secured - so it could be anything. So statements about fingerprint security being good could be "this crap again" depending on the value of the thing being secured - and inability to change it is definitely not irrelevant.

      * The blog post is too dumbed down for the title as well as the subtitle of the blog. Not that non-geeky people couldn't find it useful, but I was trapped into a boring read due to the title. Could you rename it to divenongeekystuff (geekier stuff for non-geeks) ?

      Also, not sure what you are going for with 1:365.2425, but 2 people are said to have different birthdays if they self-report with different values out of 366 possible values. Even there, it is more like 365 - the people I know with birthday 29 Feb generally self-report with a birthday of 1 Mar. I guess for fear of missing a party about 75% of years.

      --
      Bingo Dictionary - Pragmatist, n. A myopic idealist.
    3. Re:Oh FFS, not this crap again by swillden · · Score: 1

      Your blog post says security of fingerprint authentication depends on the value of the thing being secured.

      It says that as well, but the main thrust of the post is that rotation is useful within the password security model, since passwords require secrecy and rotation restores it, but that rotation does not do anything useful for biometrics, because their security -- to the degree that they have it -- is based on the integrity of the acquisition process, not secrecy.

      So statements about fingerprint security being good could be "this crap again" depending on the value of the thing being secured - and inability to change it is definitely not irrelevant.

      Sure, the security could be lousy, but the inability to change your fingerprints is irrelevant regardless. They're not secrets, they were never secrets, they never will be secrets, and there's no point in rotating non-secrets.

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
    4. Re:Oh FFS, not this crap again by bingoUV · · Score: 1

      It says that as well, but the main thrust of the post is that rotation is useful within the password security model, since passwords require secrecy and rotation restores it, but that rotation does not do anything useful for biometrics, because their security -- to the degree that they have it -- is based on the integrity of the acquisition process, not secrecy.

      Yeah, there you just hand-waved it away. The actual secret is the electronic data that goes to the machine - that is secret whatever you say and that is all that matters. Whether that is secret due to the physical machine ensuring an alive human at the other end, or due to Bruce Lee protecting the machine from attackers showing a JPEG of someone's fingerprint. And whether or not the geometry of one's fingerprint is really secret or not. No one cares.

      The data, or the process to convert a finger-print geometry into data are the real secrets here. As soon as leaked, it would be comforting to be able to change one's fingerprints.

      India is conducting the largest experiment ever in fingerprint authentication security - attacks with fixed fingerprint data have already been made, and thwarted. It is unknown whether the attackers stored the data that a fingerprint machine generates, or whether they stored the fingerprint geometry in, say, an image. They were thwarted because exact same data was used for lots of authentication requests - whereas in actual fingerprint scanning the data would subtly change every time due to different angles presented to the machine.

        It is also unknown whether there were any attacks that were not thwarted - Indian government authorities are not really paragons of transparency.

      --
      Bingo Dictionary - Pragmatist, n. A myopic idealist.
    5. Re:Oh FFS, not this crap again by TranquilVoid · · Score: 1

      Very clearly-written blog post. The final paragraph nails the problem Slashdot is having with it - context. Biometrics make more sense on a portable device with locked-down hardware and an OS that sandboxes its programs. On a generic desktop that could run any OS, and where hardware drivers could be trivially substituted, and wishing to log in to a random forum on the web, not so much. Slashdot is thinking more about the latter.

  8. Steal all the biometric files by jfdavis668 · · Score: 4, Insightful

    Instead of breaking in and stealing passwords, break in and steal all the biometric files. Your fingerprint may be unique, but to identify you I have to have a copy. If someone steals that copy, you are now compromised in a way you can't correct. You can't change your fingerprint every 60 days.

    1. Re:Steal all the biometric files by hcs_$reboot · · Score: 1

      To be fair, according to Apple, the probability that someone's else fingerprint fits yours is 1 / 10000, based on current software / technology Apple is using ; that's far from being unique. Besides, I do agree on your concern - and fingerprint is not even the worse biometric data (since it requires contact).

      --
      Slashdot, fix the reply notifications... You won't get away with it...
    2. Re: Steal all the biometric files by nehumanuscrede · · Score: 1

      I suppose you could prefix a pin or password to the biometric data which, when hashed, creates a unique signature.

      But then it takes us right back to square one in requiring a variable password or pin to begin with, so what would be the point.

      Is a trade off I guess.

      Passwords are easier to compromise, but also easier to change. Biometrics, not so much.

    3. Re:Steal all the biometric files by jfdavis668 · · Score: 1

      Not in my experience. People's fingers are not always placed on the sensor the same way. Your system needs to do an analysis to compare them that can't be done with a hashed file.

  9. Protected vs Unprotected by nehumanuscrede · · Score: 1

    While other solutions may be more effective at preventing misuse by third parties, you are not required to give your password to law enforcement without reasonable cause.

    Their simply demanding it is not " reasonable cause ".

    Whereas your biometric ID is fair game. They can, and have, walked into an establishment and forced everyone who used biometric fingerprints to unlock their phones to do so. You have no recourse.

    I'll keep my passwords until they fix the other problem thanks.

    1. Re:Protected vs Unprotected by Known+Nutter · · Score: 1

      They can, and have, walked into an establishment and forced everyone who used biometric fingerprints to unlock their phones to do so. You have no recourse.

      Citation needed. That sounds ultra-paranoid.

      --
      Beware of the Leopard.
    2. Re:Protected vs Unprotected by tepples · · Score: 1

      They can, and have, walked into an establishment and forced everyone who used biometric fingerprints to unlock their phones to do so.

      What's the difference between that and "enter your password while I look away"?

    3. Re:Protected vs Unprotected by Known+Nutter · · Score: 1

      http://bgr.com/2017/10/16/touch-id-unlock-cops-search-warrant-illinois/

      That's nice, but it does not begin to cite the GP's assertion that "they [whoever they is] can, and have, walked into an establishment and forced everyone who used biometric fingerprints to unlock their phones to do so. "

      --
      Beware of the Leopard.
    4. Re:Protected vs Unprotected by DCFusor · · Score: 1

      It's not working for this guy: https://www.theregister.co.uk/... and some others who have been jailed for contempt for refusing. Funny thing about courts - your interpretation of a law means nothing to a judge. Only his matters.

      --
      Why guess when you can know? Measure!
    5. Re: Protected vs Unprotected by sysrammer · · Score: 1

      That *is* my password, you insensitive clod!

      --
      His ignorance covered the whole earth like a blanket, and there was hardly a hole in it anywhere. - Mark Twain
  10. because.. by epyT-R · · Score: 4, Informative

    1. They aren't tied to biometrics, which once compromised, aren't easily changed. Plus, many people find it instinctively invasive, possibly because of that reason. In contrast, passwords/x509 are easily changed when when compromised or forgotten.

    2. Biometrics work as authenticators but not as authorizers.. Nothing stops someone from duplicating your biometric properties (pic of your fingerprints or irises/face) without your authorization. Not so with a password.

    1. Re: because.. by Corbets · · Score: 1

      What you meant was âoeidentifiers, not authenticatesâ. Gotta understand it if youâ(TM)re gonna argue it.

    2. Re: because.. by epyT-R · · Score: 1

      Sorry, I stand corrected.

  11. Longeivity by JBMcB · · Score: 1

    Algorithms to validate, store and process passwords have been around a LONG time. Best practices are well known, and are relatively simple. You can build a password-based access control system using off the shelf libraries and known patterns that is very difficult, if not impossible, to bypass. The limiting factor to it's success is human fallibility.

    Nearly everything else is complicated, involves a lot of math that not a lot of people understand, or third party hardware you might not trust, or third party services you might not trust, etc.. etc.. etc...

    On top of all of that, maybe you can mitigate *some* human fallibility, but it can still come into play.

    --
    My Other Computer Is A Data General Nova III.
  12. Yubikey by darkain · · Score: 1

    Just support devices like Yubikeys everywhere. Done.

    This is what I use for Google/Gmail, Facebook, Github, and anything that requires SSH access. No more passwords. Just a physical device with a simple pin code.

  13. The answers by Okian+Warrior · · Score: 4, Interesting

    The answers are pretty obvious.

    Firstly, we still use them because there's no reasonable replacement. Duh.

    Secondly, there's no reasonable replacement because of the way our computers work.

    Passwords are essentially information held in a system outside the computer (your head), that can be used for verification. The problem is that humans aren't really good at remembering passwords, and we need so many of them, and they are infrequently needed.

    All attempts at using computers to solve this issue have run afoul of the "general purpose computer" problem: because our computers do not address security properly, we cannot guarantee what software is running on the local hardware. We cannot guarantee the security of passwords held on the computer, or in an encrypted file, because it's so easy to download and run malware. No one keeps track of all the things run on the computer, and we can't even trust the people who supposedly *do* keep track.

    One reasonable solution is to use hardware specific to the purpose that's *not* a general purpose computer.

    If you had a piece of hardware - a thumb drive, for example - that was *not* general purpose and could not download and execute code, then that could be made pretty secure. It could hold a person's private key, have functions to encrypt, decrypt, and sign documents, and also pass out the public key. It could also download and install new keys, with the understanding that the base functions could not be changed.

    There's some details involved: you need a way to securely backup the data, and you need a way to securely recover the data in various situations. Mostly, you need to save the data somewhere safe and write down a master password (one, a PIN of sorts) somewhere else.

    The Mooltipass is pretty close. It generates strong passwords for each web site registration, and will fill in the fields for you when you go to log in.

    That's not the complete solution, however. It should *encrypt* the password with the user's private key and the site's public key so that no one can view it(*), or even better use a zero-knowledge authentication process.

    If we could somehow begin using a fixed-program computer - say, something the size of a credit-card calculator that requires a pin and that holds the information for *all* the cards in your wallet - we could get away from passwords.

    We would also have a single point on which we could put *all* our effort to make secure.

    Hypothetically, that one card would reduce credit card fraud to near zero. When you use the card you enter your PIN on the keypad, and the card generates a ShopSafe number tied to your credit account, valid for one purchase.

    Take a look at the badges at high-tech conferences these days. It seems like the hardware shouldn't be that hard or expensive.

    Could this be the next killer product from Apple? A hand-held thingy that's secure and ultra-convenient, that you use for payments (IRL and online) and password entry?

    (*) Yes, ssh is not absolutely secure. Did you think all those cert authorities in your browser have been properly vetted?

    1. Re:The answers by duke_cheetah2003 · · Score: 1

      I don't like the hardware answer to this problem. It's fraught with issues. The biggest one, which I encountered personally, is what happens when your device is irrecoverably lost. Or in my case, my phone number was lost. I could not get the same phone number, I had to get a new one, and it made fixing sites I had doing 2FA via SMS to my phone number a real bitch to recover. Took many emails and a sob story to get 2FA stripped from my account so I could get back in.

      This is a serious problem. If Google stuck to their guns and refused to dissociate my unrecoverable phone number from my account, I'd be shit out of luck. Google should have stuck to their guns, but they didn't and I recovered my account.

      What happens when this happens and the sites you've lost access to aren't going to be helpful? You're screwed. This is why hardware devices are a non-solution. They create more problems than they solve.

      I don't have a solution, but I know this, the solution has to be something that has all the features of a password:
          1. You make it.
          2. You store it in your mind.
          3. You can change it at any time.

      Any solution that doesn't have those three features is non-starter. It must have those features or it's worthless, to me at least.

    2. Re:The answers by Kjella · · Score: 1

      Except that a dongle is equally unreliable at telling you what you're agreeing to, which is pretty essential to signing otherwise it's purely an authorization token.

      --
      Live today, because you never know what tomorrow brings
  14. Sometimes the first impulse is right by goombah99 · · Score: 5, Insightful

    think of them as a mutable biometric. it's biometric because its stored in your brain. It's mutable because you can change it. it can't actually be stolen from you if you don't give it up or write it down.

    it's only when you go to transmit it that the problem occurs.

    When you look at this this way, then you see that things like finger prints or retina have the same problems and worse. they are not mutable, they can be taken from you without you knowing it, and the transmission layer is still vulnerable

    Nearly always, your first solution to a problem is the best one. Not always of course or there would be no need to research and study. But people have been using passwords for milennia because they are an effective tool that works from giving something to the sentry, to logging into google.

    --
    Some drink at the fountain of knowledge. Others just gargle.
    1. Re:Sometimes the first impulse is right by lhowaf · · Score: 2

      I prefer passwords, too, but they can be taken from you without your knowledge. The problem is passwords (or hashes) are stored at the places you authenticate. Even when you use unique passwords, there is a potential breach for each site/authentication pair and there's nothing you can do about that. That said, the point about biometrics not being mutable while still being vulnerable to loss is key.

    2. Re:Sometimes the first impulse is right by ShanghaiBill · · Score: 1

      The problem is passwords (or hashes) are stored at the places you authenticate.

      You say that like storing the password and storing the hash are somehow equivalent. They are completely different. Passwords should never be stored. But storing salted hashes is standard practice, and is secure for modern hashes, especially when combined with limited attempts and credential verification.

    3. Re:Sometimes the first impulse is right by Altrag · · Score: 1

      Biometrics have the same problem. Once your face scan is taken, its just a bunch of ones and zeros same as your password is, and is subject to all of the same storage and security requirements in order to keep it safe.

      And most importantly, it wouldn't be excessively hard to create a dummy device that emulates the biometric scanner's output once you know the expected input data.

    4. Re:Sometimes the first impulse is right by swillden · · Score: 1

      When you look at this this way, then you see that things like finger prints or retina have the same problems and worse. they are not mutable, they can be taken from you without you knowing it, and the transmission layer is still vulnerable

      Mutability doesn't matter for biometrics, and neither does the fact that copies can be taken without your knowledge, because those don't affect the biometric security model. I wrote a detailed analysis here: http://divegeekstuff.blogspot....

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
  15. SLASHDOT: FIX YOUR CODE MANGLING!!! by goombah99 · · Score: 4, Informative

    people who post to slashdot from iphones and such get all of their apostrophes turned into å(TM)t â(TM)t

    THis is 2017, it's possible to parse plain text and unicode correctly now I have read.

    --
    Some drink at the fountain of knowledge. Others just gargle.
    1. Re: SLASHDOT: FIX YOUR CODE MANGLING!!! by MachineShedFred · · Score: 2

      No, sorry; it is the fault of Slashdot for still not supporting Unicode 20 years later.

      Rather pathetic, really.

      --
      Slashdot still doesnâ(TM)t support Unicode after it was added to the HTML standard in 1997.
    2. Re: SLASHDOT: FIX YOUR CODE MANGLING!!! by iNaya · · Score: 1

      Last time I checked, directional quotes were part of the English language.

      --
      The Unicode standard is over 20 years old. Why does Slashdot not support it?
    3. Re: SLASHDOT: FIX YOUR CODE MANGLING!!! by MachineShedFred · · Score: 1

      I present to you RFC 2070 which extends the HTML 2.0 spec to use Unicode, and has since been included in HTML 3, 4, and 5.

      In January of 1997 that happened.

      This is a web site, is it not? It renders HTML, does it not? Have they not actually developed anything on it since 1997? Or are they just carrying around 20 years of technical debt because of stupid excuses like "its [sic] an english language site"?

      Also, the apostrophe is a punctuation mark in the English language.

      --
      Slashdot still doesnâ(TM)t support Unicode after it was added to the HTML standard in 1997.
  16. Biometrics are ID, not security by phantomfive · · Score: 1

    Such technologies -- from fingerprint scans to facial and retinal scans -- promise more secure and reliable factors than alphanumeric passwords, the executives agreed.

    No, no no, my god, no. Something that can be acquired just by looking at you is not secure. Using as authentication something that can only be changed by destructive surgery is not sane.

    --
    "First they came for the slanderers and i said nothing."
  17. Asked and answered by petes_PoV · · Score: 1
    The article answers its own question. Why do we still use passwords? and then tells us, of the alternatives,

    the obstacles to using them are often prohibitive

    Which makes the article rather pointless.

    However it misses out a vital aspect. No matter what technology replaces passwords, it will get hacked, faked, or discovered. One day. And that means that whatever security measure is in place, it must be changeable by the user, just like passwords are.

    So that rules out all the biometric options, if they were only to be used on their own. Consequently, whatever replacement is to succeed must be something the user has (and could change) or knows (and can change).

    But what?

    --
    politicians are like babies' nappies: they should both be changed regularly and for the same reasons
  18. Don't kill them, implement 2FA+ by Murdoch5 · · Score: 1

    There is nothing wrong with passwords as a first line of authorization, but if it's all you're using then you really deserve to be hacked. In 2017 it's no longer acceptable to have a single factor of authentication to a system, especially with the prevalence of TOTP and Hardware key, such as YubiKey.

    When trying to secure servers, if you don't have 2FA+ enabled, then you should be fired and blacklisted!

    1. Re:Don't kill them, implement 2FA+ by tepples · · Score: 1

      In 2017 it's no longer acceptable to have a single factor of authentication to a system, especially with the prevalence of TOTP and Hardware key, such as YubiKey.

      Which is why I find reliance by Google and Twitter on SMS as the primary second factor, with TOTP and YubiKey relegated to backup second factors, to be unacceptable. What would you recommend for working around this unacceptable situation?

  19. When princes misbehave by Shogun37 · · Score: 1

    Sure, it's easy to say that passwords are a bad security method. But, it's like saying we need a better lock, when the walls are so thin bad guys just walk right through. Set the system up right, maintain it, and change when needed. At that point, the damage from easily guessed passwords is minimal.

  20. Services still require a mobile phone number by tepples · · Score: 1

    But have you figured out how to U2F key with Google or Twitter without first setting up mobile phone verification? Say I want to have U2F (such as YubiKey) as my primary second factor, with TOTP (such as Google Authenticator) as a backup. But services like Google and Twitter support these only as backup second factors, not primary second factors. If I try to set up one of those as a second factor on Google or Twitter, the site won't let me proceed past the mobile phone verification. I don't want to use a mobile phone as the second factor for two reasons:

    Cost U.S. pay-as-you-go carriers charge 10 cents per received text message, and services like Twitter automatically send the code as a text message to the associated mobile phone even if I have a non-SMS second factor set up. SIM swap fraud SMS authentication is vulnerable to social engineering in which the attacker compromises an account by arranging delivery of a replacement SIM to him.
    1. Re:Services still require a mobile phone number by q4Fry · · Score: 1

      In case it is helpful, I griped about this, too, and swillden responded, suggesting that I remove the phone number after adding another 2FA.

    2. Re:Services still require a mobile phone number by tepples · · Score: 1

      From swillden's reply: "Use a payphone, if you can find one." I don't see how that would work, as a payphone can't receive text messages, and Twitter can't verify a phone number through a voice call.

  21. Disposable personal authentication devices by vakuona · · Score: 1

    Here is a thought. How about multipurpose disposable personal authentication devices.

    Think of TouchID. They key thing about TouchID is that the biometric authentication is "on device". So if you decoupled the TouchID from the iPhone, and developed a token that could use generate a one time passphrase that you use to login to any website, that would mean an attacker needs physical proximity to you to steal your logins. Goodbye Russian hackers.

    Single point of failure yes, but also single point of hardening.

    1. Re:Disposable personal authentication devices by Average · · Score: 1

      Welcome to the world of FIDO U2F (fidoalliance.org). The best-selling U2F device these days (the YubiKey line... note that FIDO U2F is only one mode those work in) does not do biometric authentication before responding with one-time public-key-based security, but there are more expensive U2F devices out there already that do local biometric unlock.

    2. Re:Disposable personal authentication devices by DCFusor · · Score: 1

      I'd suggest googling yubikey weakness and looking at the bug list before trusting one. Or any hardware/software device you don't really control the innards of.

      --
      Why guess when you can know? Measure!
    3. Re:Disposable personal authentication devices by John.Banister · · Score: 1

      It would be nice to see a device that combines the features of YuiKey 4 Nano with Kensington VeriMark. My fingerprint wouldn't get stolen, because it wouldn't leave the device, but the device wouldn't provide authentication without both my fingerprint and a password. I suppose a little more security in public places could be added by the device also requiring the presence of an RFID keyfob I'm not seen to be using.

  22. control by Bobtree · · Score: 1

    Because all the big sites wanted to be OpenID providers but not to accept logins from elsewhere.

  23. and the biometrics can change by swschrad · · Score: 4, Insightful

    accident damage, surgery, degredation of the eye are some of the ways you can be locked out of a biometric identifier. as the population ages, this is an issue that you need to think about. I will not use bios for this reason, as for some reason, I am not getting younger and more invulnerable.

    --
    if this is supposed to be a new economy, how come they still want my old fashioned money?
    1. Re: and the biometrics can change by Monster_user · · Score: 1

      Weight gain or loss, and aging, affects most other biometrics. I wouldn't want to rely on a system which I where I couldn't retain, or might have difficulty maintaining, the authentication "credential".

    2. Re: and the biometrics can change by Anonymous Coward · · Score: 4, Insightful

      Passwords are âoesomething you knowâ, and would be impossible to hack, or obtain, since itâ(TM)s metaphysical and in your brain. The trouble with passwords is inconsistency and bad practice.

      Expiring passwords was recently deemed a bad idea by NIST, since it leads users to simple patterns and incremental changes of least-complexity allowed.

      Furthermore, any system which allows arbitrarily large guesses is weak, not the password itself. It doesnâ(TM)t matter if itâ(TM)s biometric, passwords, tokens, sms messages, anything. All these systems send binary numbers representing the measured input, and I can code a computer to log in using random or incremental data. With enough guesses, Iâ(TM)m going to be right. The only defense is a system where wrong guesses locks accounts, and a different form of identification is required. The iPhone does this best IMHO. You can use your finger only a half dozen times incorrectly before it moves to passcode. 10 wrong tries with that, and wipe. Go back to your backups.

      Itâ(TM)s astonishing how well it works, contrasted with how poorly the rest of Appleâ(TM)s authentication mechanisms work.

      The other problem is systems which store the plain-text of passwords. When a site gets hacked with users and passwords, they werenâ(TM)t using a one-way hash, with a good user-unique salt, and the passwords are now compromised. Since uses tend to repeat a password over multiple services, your password is now only as strong as the weakest system. This is where two-factor comes in to play. A few wrong two-factors and the account should get locked, and the password should be expired.

      Good systems will lock after too many bad tries, and passwords should be two-factored so reused passwords that were exposed online are rendered ineffective.

    3. Re:and the biometrics can change by david_thornley · · Score: 1

      Also, some people don't register very well. Fingerprint scanners are pretty bad at recognizing me, for some reason. I turned off fingerprint identification on my iPhone because it almost never worked, and if it wasn't going to help me I didn't want it suddenly deciding to work when someone forces my finger on the sensor.

      --
      "When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
  24. A way better qyestion is by bobstreo · · Score: 1

    Why a 4 digit pin code is considerd secure?

    Passwords for something you care enough about to protect are only the start. Businesses have been using TFA either Secure ID or via text for years.

  25. Nothing better. by gurps_npc · · Score: 1

    Biometrics make horrible passwords. They are way too easy to steal and copy. Two factor makes more sense, but it is still a password. I suggest a many-factor system.

    What we do in real life is use a combination of multiple different methods. It's not just the way your face looks, but the location, clothing, and voice.

    A complex system that combines multiple methods, assigns a percent sure of identity might work. It could include a simple password (six characters), that must be typed on live video (increasing both the time it takes to enter a password and allowing verification of keyboard and typing pattern along with your hands and video file is hashed and saved to be sure it isn't reused), along with a check for an existing cookie on the PC, with a verification for usual browser, usual source IP address, etc.

    Such a system would be set up to allow for changes, but not all at once. I.E. If you just change your keyboard, that doesn't trigger a violation, but if you replace your computer then you better have the same router source IP address or expect a phone call/email/letter.

    --
    excitingthingstodo.blogspot.com
  26. Yes, I'm that guy by konohitowa · · Score: 1

    It's Akamai, not Akamia.

  27. The best safe by duke_cheetah2003 · · Score: 1

    Alas, as of this moment, the one place you can put something that NO ONE else can get it is in your mind. This is the ultimate safe.

    This is why passwords remain the preferred authentication method. Because it feels secure, your mind is the only place the key exists. As soon as you move that key out of a person's mind and into a device, or biometric, it's no longer in the best safe in the world. Your mind. It's a very important semantic. People feel passwords are safe because they're stored in the best safe. Your mind.

    Until we come up with another way to store the key in your mind, password will reign king of authentication. Period.

    Personally I don't really like biometrics, or 'devices' to store your passwords for these exact reasons. Parts of my body can be taken from me by force. My 'device' can be lost, stolen, or damaged, locking me out of everything. The mind is insurmountable security. Yes, you can beat someone over the head to try to convince them to spill the beans, but even that can be resisted by a determined person.

    The password is still the best security device we have. It can't be lost, stolen or damaged. No one can 'steal' it from you (though it can be guessed by persistent actors, but so could any other authentication method. Just takes time and effort.) It's no going anywhere until a suitable replacement is created.

    And that suitable replacement doesn't seem to be coming anytime soon, because nothing feels like a password. It's yours, you made it, and you can change it. You can't change your fingerprint or retina scan.

  28. Requires Chrome; git push still needs password by tepples · · Score: 1

    [YubiKey] is what I use for Google/Gmail, Facebook, Github

    How does that work?

    As far as I can tell, U2F on GitHub is incompatible with Mozilla Firefox, incompatible with Edge, and incompatible with Safari. I'm not even sure it works with other Chromium forks, as the page mentions Chrome. In addition, you need to buy a supported smartphone or tablet first because U2F requires working TOTP, and you still need to generate a password for use when pushing.

  29. Biometrics are provably not secure by Antique+Geekmeister · · Score: 1

    Fingerprints are easily forged. The excellent paper http://web.mit.edu/6.857/OldSt... covered the issue 15 years ago and remains valid with even the best modern fingerprint scanners.

  30. biometric binary blobs by epine · · Score: 1

    A person determined to use passwords in a sane way (every password unique, with 60+ bits of true entropy) enjoys at least a modicum of confidence that the password implementation itself is simple enough to actually work as implied.

    I'm about fifty years away from believing than any biometric security solution can be trusted without inspection (we still need some astounding advances in proof-of-correctness technology).

    And I don't really feel like reading all that code, anyway. Theo and his crowd probably won't do it for me, on principle.

    Every problem in computer science can be solved by adding another layer of indirection, except for too many layers of indirection.

    Every problem in computer security can be solved by adding another trust authority, except for the proliferation of trust authorities you already have no compelling reason to trust.

  31. "The question is what to replace them with." by QuietLagoon · · Score: 3, Insightful

    Bingo! Biometrics suck. How do you change your fingerprints, or your eye's iris?

    1. Re:"The question is what to replace them with." by 93+Escort+Wagon · · Score: 1

      How do you change your fingerprints, or your eye's iris?

      They do it in movies and anime all the time.

      --
      #DeleteChrome
    2. Re:"The question is what to replace them with." by swillden · · Score: 1

      Bingo! Biometrics suck. How do you change your fingerprints, or your eye's iris?

      This is irrelevant. Your erroneous argument arises from the mistaken application of the secrecy-based security model of passwords to the acquisition process-based security model of biometrics. I explained at length here: http://divegeekstuff.blogspot....

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
    3. Re:"The question is what to replace them with." by houghi · · Score: 1

      Bingo! Biometrics suck. How do you change your fingerprints, or your eye's iris?With a fork.

      --
      Don't fight for your country, if your country does not fight for you.
  32. Something you know, ... by thegarbz · · Score: 1

    There are three factors for authentication. Something you know, something you have and something you are.

    Why would we give up one in favour of another when we could adopt the radical idea of using TWO AT ONCE.

  33. Passwords have the same problem as credit card #s by Solandri · · Score: 1

    They're unique, but stay the same between uses. So if someone manages to copy it when you use it, they can use the copy in the future to pose as you.

    Fortunately, that means they have the same solution as credit cards. Chip and pin works by you remembering a PIN (like how you remember a password). You enter the PIN into an authorized device, and that allows the device to query the chip. The chip then establishes a secure link to the processing site. Intercepting that session's communications doesn't make it any easier to forge a future communication.

    Likewise, passwords can be replaced by a authenticator. Your password unlocks the authenticator. The authenticator then takes the site you're trying to login to and the time of day to generate a unique code you need to login. That way your password never has to leave your control. In theory this could be used in lieu of a password, but so far it's mostly being used to augment your password. That is, you still use a password (which can be stolen) to login to the site, but you also need the authentication code as a second factor to let you in.

    This is mainly because Google's implementation is half-assed and lets you use it if you have access to the device (which is always for phones without security enabled. Authy is better implemented, requiring a passcode or password to use every time, backs up your authentication keys on the cloud so you can share them between multiple devices (they're still useless without a passcode/password), and is compatible with Google Authenticator. It's still vulnerable to some sort of keylogger. So ideally, this authenticator would be a separate physical device which did only authentication so there's no opportunity for rogue software to be installed onto it.

  34. Biometrics more secure? by OneHundredAndTen · · Score: 1

    It has been shown, time and again, that biometrics can be beat (and are beat) by relatively low-tech approaches - sometimes very low-tech approaches. And, to add insult to injury, once compromised, biometrics cannot easily be revoked, if at all. Use biometrics at your own peril.

  35. Biometrics are riskier by bingbong · · Score: 1

    You can change your password, but you canâ(TM)t change your fingerprint.

    Aside from all the low tech ways to defeat biometrics (gummy bears anyone?), the simple issue is if your biometric information gets compromised, youâ(TM)re toast.

    --
    "Omnis tuus capsa sunt inesse nos"
  36. Solution is there if you care by Karmashock · · Score: 1

    https://en.wikipedia.org/wiki/...

    This or something like it... fin.

    --
    I've decided to stop wasting my time responding to AC trolls/sockpuppets... so if you want a response from me... login.
  37. Password change=surgery by Roger+W+Moore · · Score: 1

    The gradual change of your biometrics over time is the least problematical of the issues with biometrics. While this is annoying it is easy to fix by rescanning the information after first proving who you are by some other, manual, means.

    The problem with biometrics is that if my information gets hacked the only way I can change it is via surgery and I'm simply not willing to have eye surgery to change my iris if my iris pattern is hacked when I can change a password simply by thinking of a new one.

  38. Stop using passwords.... by modmans2ndcoming · · Score: 1

    Start using Biometrics AND Public Keys. Multi Factor is always better than single factor. PKI can be convenient if the syustems are in place to use them. Imagine world where you use your fingerprint and a Public Key to get access. THen you can generate public keys for every transation you make. Finger print to prove you are present, the system sends a message signed with the key you gave it, you decrypt using your private key and send a reply back with the random data in the message (signed with their public key of course) to show that you are authorized. Even changing our credit system to the later hald would be 1000 times better than SSN.

  39. Identification v. Authorization by Macdude · · Score: 1

    Why can so many people, even people "in the industry" not understand the difference between Identification and Authorization.

    Biometrics is a good form of Identification, it's hard to lose your fingerprint or your retina (it can happen but it's not common in everyday life). You can't forget them at home, your spouse can't take yours with them by mistake, etc. A biometric ID/Authorization system can be excellent, near perfect if fact, at identifying you as you but it has no ability to handle the situation when it identifies someone who isn't you as you -- it has no system to handle the false positive.

    Biometrics are a lousy form of Authorization. Once your biometrics have been used to identify you, you need a separate system to authenticate that it's actually you -- to defeat the false positive. As you travel around you leave your fingerprints and DNA all over the place, your image is recorded hundreds if not thousands of times a day. Your biometric ID is not private, gathering the information required to impersonate you is easy even if the techniques to impersonate you are not readily available at this time.

    There is also the issue of what to do if your biometric identification has been compromised. If someone is accessing your bank accounts because they've been able to successfully fake your fingerprints, the bank can't issue you new fingerprints.

    Biometrics are great for identification, but are terrible for authorization.

    --
    "Grab them by the pussy" -- President of the United States of America
  40. Re:Simple... by myowntrueself · · Score: 1

    Text input is the one universal constant we have for communicating between user and computer.

    And requires only the same basic hardware that is required for general purpose user input...

    --
    In the free world the media isn't government run; the government is media run.
  41. Because by markdavis · · Score: 4, Insightful

    >"Why Are We Still Using Passwords? "

    Because they are cheap, generally convenient, proven, and understood. Passwords actually work quite well *IF* they are managed correctly. And despite the summary, dictionary attacks are generally useless when servers are configured correctly.

    For high security, when necessary, combining a password with a token of some sort is extremely effective.

  42. Nice try by Godwin+O'Hitler · · Score: 1

    Now fuck off trying to fool people into making your job easier.

    --
    No, your children are not the special ones. Nor are your pets.
  43. Biometrics are not secure in the USA by Anonymous Coward · · Score: 1

    The problem with using biometrics is the US courts can compel biometrics. They are not protected the same way a password is. There is case law supporting this, so until that is fixed I'll stick with a complicated password, and encryption.

  44. Biometrics are only local authentication by FeelGood314 · · Score: 1

    They only authenticate you to the machine reading you. I can't use biometrics on line unless the machine reading me is already trusted. So how does a bank trust the finger print scanner?

    A secret is always going to be the best security. However, how knowledge of the secret is verified can can be improved in a lot of ways.

  45. Because passwords are the best by smblion · · Score: 1

    because passwords are best.

    You disagree? Invent a new method of opening the door to your house or starting your car before suggesting your "amazing new idea".

  46. A Fingerprint is a Password by OYAHHH · · Score: 1

    And for most of us a fairly permanent one at that.

    What an absolutely asinine statement by "the executives".

    --
    Caution: Contents under pressure
  47. Re:my password is "ThePopeIsPedophile" by sysrammer · · Score: 1

    Who is Popels?

    --
    His ignorance covered the whole earth like a blanket, and there was hardly a hole in it anywhere. - Mark Twain
  48. Question and Answer are provided by srichard25 · · Score: 1

    The question posted is: Why Are We Still Using Passwords?

    The answer is provided in the summary: Stronger and more reliable alternatives to passwords already exist, but the obstacles to using them are often prohibitive

    Nothing more to say.

  49. Re:Biometricsare not secure by softcoder · · Score: 1

    How do you propose to log into a website with your fingerprint? Put it on the reader and send it to the website? Unencrypted? Oh Oh now instead of just a password compromised, if it is intercepted, your fingerprint is out there for anyone to use
    And what makes you think the websites will be any more diligent about safeguarding your fingerprint (or the encrypted version thereof) than they are about safeguarding your password (or an encrypted version thereof)?

    Using biometrics as a 'login' device is an insanely crazy idea. The only reason people are sort of accepting it is because the flaws have not been exposed by the constant attacks of millions of cyber criminals.
    Once your digitized fingerprint or face has been stolen you are in a world of trouble, worse than if your Equifax data were stolen.

  50. Re:Biometricsare not secure by dmr001 · · Score: 1
    I have plenty of apps on my phone that are essentially websites, that, one I verify my identification by other means (like my password and some other factor like my pre-registered IMEI number or out of band code sent to me) let me log in with my fingerprint. Which isn't transmitted; the phone has an API that tells the app my fingerprint was recognized.

    This includes my bank, investment firms, and hospital (that's the one keyed to my specific phone).

  51. Client side TLS certificates by u801e · · Score: 1

    When you set up your account, you generate your certificate signing requests for each device you plan to use and send them to service which will verify your identity and sign your certificates. Then you configure your browser to use the certificate when you connect to the website. That will be how you authenticate.

  52. Legal protection by Machupo · · Score: 2

    A password provides you legal protection from being (legally) forced to divulge it, not so with biometrics or hardware authentication dongles.

    Multifactor is always better, but a key component of that has to be something hidden in your mind.

    --
    *insert pithy sig here*
  53. Anything biological is not an alternative by Anonymous Coward · · Score: 1

    To prevent fraud and to protect privacy, credentials must be separate from people's physical and legal identity. No biometric key can prevent coercion, guarantee anonymity or be revoked when it is compromised.

  54. Ease of use by Anonymous Coward · · Score: 1

    Right now, I can go to pretty much any computer in any country and be able to log into gmail (for example) without having to have:

    a) elevated privileges to use/add additional hardware, OR
    b) a mobile phone, OR
    c) loose scraps of paper with one-time pads, OR
    d) anything similar

    Granted if I'm at some dodgy internet cafe there is a risk of key loggers, but passwords can be changed when you get to a trusted location.

  55. Duplicate Post by zifn4b · · Score: 1

    https://ask.slashdot.org/story...
    https://it.slashdot.org/story/...
    Let's re-hash the same old crap and get advertising revenue, yay.

    --
    We'll make great pets
  56. Outsource authentication by StickyKeys · · Score: 1

    I never store passwords if I develop a UI. Either I use single-sign on so authentication is handled by someone else. If that's not feasible I just force users to verify their email address if they want to login, typically email address verification takes less than 15 seconds and the user will only need to do this once per device.

  57. Because that's what we should always use by Anonymous Coward · · Score: 1

    Passwords cannot be chopped off and used while you aren't there. Passwords can be invented and changed at any time by the user. Their very nature makes them ideal for security: an intangible, boundless bank of secrets known only to you that can be changed any time. Until they come up with some other identification method that doesn't rely on tangible objects or physical attributes that cannot be changed, the password is where it's at.

  58. Biometricas are resused, unchangable passwords. by Blue23 · · Score: 1

    Until the biometric device is talking directly without any middlemen (like the vendor or the internet) to the payment people, it is inherently less secure. Because at all the points between, it's just a digital password, and one that is (a) reused between sites and (b) unable to be changed.

    --
    LITTLE GIRL: But which cookie will you eat FIRST? C. MONSTER: Me think you have misconception of cookie-eating process.
  59. Re:Biometricsare not secure by bingoUV · · Score: 1

    So you authenticate to your phone, and your phone authenticates to the "app", which is a website. An unauthorized person needs to break only one of these authentications. This approach doubles the variety of the possible attack vectors.

    --
    Bingo Dictionary - Pragmatist, n. A myopic idealist.
  60. Public keys by MikeBabcock · · Score: 1

    Public key cryptography has been the solution for a long time, we've just shot ourselves in the feet by not implementing it on the client side.

    Client-side certificates for instance, SSH keys, PGP/GPG keys -- they're all examples of public key cryptography that works to secure data and requires the user to only remember one good password instead of many.

    --
    - Michael T. Babcock (Yes, I blog)