Tech Broiler

Jason Perlow and Scott Raymond

Google, Facebook: End Passwords, Get Biometrics. Now!

By | February 4, 2011, 10:25am PST

Summary: With the multitudes of accounts we have to deal with for email, social networking and other applications that require password authentication, we need a better solution.

With the multitudes of accounts we have to deal with for email, social networking and other applications that require password authentication, the risk of compromise is too great. We need a better solution.

So this morning I did the usual. I woke up, got out of bed, I answered the call to nature, I popped a K-Cup in my Keurig brewer, and I shuffled downstairs to my home office and logged into my personal email account.

This is the first thing that I saw:

Needless to say, I was not amused. At all.

Now, I generally regard myself as extremely careful with my computer security. To the point of being extremely paranoid about it. I use “strong” passwords, mixed alphanumerics with non-alpha characters. An example of this would be something like R1tch13R1c4386!

Not only that, but I don’t use the same password on all my services. My Google password is unique.

Today, as modern computing users, we’re inundated with passwords on all sorts on web and social networking sites. I use GMail and all the Google Apps, such as Calendar, Analytics, Docs, et cetera. I use FaceBook. I use LinkedIn. I use two separate blogging accounts, and I have logins on a  myriad of other websites and web-based applications, not to mention all the corporate intranet stuff I deal with on a daily basis.

It’s gotten out of control. Keeping track of these requires spreadsheets and documents, stored in various places, because you can’t possibly hope to remember them all and when they expire. And then of course you need to have them reset all the time with your new temporaries sent into your email should you forget them.

So back to my GMail account. Someone had clearly compromised it, this despite the fact that I use strong passwords. Not only do I use strong passwords, but I use the Chrome browser as my standard on all my PCs, no matter what OS I use, arguably the most secure browser available today, and it’s the only one I access the GMail web interface with, using an encrypted connection.

I also use Linux as my primary operating system, with my Windows software running virtualized, each instance with antimalware and antivirus software running on them, and I don’t run Internet-facing web apps from those. My corporate applications and email are isolated in a Virtual Machine using an encrypted virtual hard drive.

My PCs aren’t the only devices that talk to my Google account. I have two Android phones, as well as an iPad. So the attack vector could have been from anywhere.

With all of those precautions in place on my PCs, I have no idea how that account was compromised. I can only speculate: It could have been on a rogue Android or iOS app, it could have been a cross site authentication thing on FaceBook, or it could have been as something simple as a email or web-based phishing attack, although I tend to be pretty vigilant about obvious phishing emails which come across my desk on a daily basis now.

It could also have been a “Brute Force” attack, although with “Strong” passwords that becomes more difficult. I also won’t rule out Google’s servers being penetrated directly, although that seems less likely.

The point is, it doesn’t matter. If someone like me can get compromised, so can anyone else, especially someone who isn’t keeping track of their online accounts and behavior as much as I do.

Let’s face it — passwords suck. Once someone knows what they are, your security is in a world of poo. I would have used a much stronger term than “poo”, but I’ll let Private Pyle do this for me.

There is a better solution than passwords. That solution is Biometrics.

[Next: Eliminating Passwords with Biometric Devices]»

Topics

Jason Perlow, Sr. Technology Editor at ZDNet, is a technologist with over two decades of experience integrating large heterogeneous multi-vendor computing environments in Fortune 500 companies.

Disclosure

Jason Perlow

My Full-Time Employer is IBM. I write as a freelancer for ZDNet.

Disclaimer: The postings and opinions on this blog are my own and don't necessarily represent IBM's positions, strategies or opinions.

I own no investments or direct financial instruments in the companies I write about.

Biography

Jason Perlow

Jason Perlow, Sr. Technology Editor at ZDNet is a technologist with over two decades of experience with integrating large heterogeneous multi-vendor computing environments in Fortune 500 companies. A long-time computer enthusiast starting the age of 13 with his first Apple ][ personal computer, he began his freelance writing career starting at ZD Sm@rt Reseller in 1996 and has since authored numerous guest columns for ZDNet Enterprise and Ziff-Davis Internet. Jason was previously Senior Technology Editor for Linux Magazine, where he wrote about Open Source issues from 1999 to 2008.

In his spare time, Jason is an avid amateur chef and food writer, where his work reviewing New Jersey restaurants has appeared in The New York Times. He is also the founder of the popular food web site eGullet and blogs about restaurants and cooking at OffTheBroiler.com.

75
Comments

Join the conversation!

Just In

RE: Google, Facebook: End Passwords, Get Biometrics. Now!
JACOBSONR 14th Oct
Good day to confirm this comment I would appreciate T h e b e s t o f Z D N e t d e l i v e r e d your website very nice to everyone Yes, Oracle is the only one with shared-disk architecture, but that is there advantage. It means you can add or remove nodes and the database lives on. In a shared nothing architecture, if you lose a node, you lose the system. I'm sure Oracle appreciates EMC highlighting their advantage.I also desire to signal in your RSS feeds. Thank you as soon as once again and maintain up the great operate Awesome post! Thank you very much || thanks for nice content this is really benefit to me.
0 Votes
+ -
That won't work
FADS_z 4th Feb 2011
Even your pc has finger-printer scanner, how about your phone? How about going to library pc and access your account?

Some centralized authentication may be the answer. The problem is who. Can we let goverment handles our account. Microsoft tried this a couple years ago, but few trusted them. Now we have this dilemma.

Right now I am using one small applicaton I wrote myself. All username & password saved in xml file and encrypted, clicking the url link will bring me directly into login page, so I don't need remeber url too.
@FADS_z
Sounds similar to LastPass.
0 Votes
+ -
RE: Google, Facebook: End Passwords, Get Biometrics. Now!
SecurityThroughObscurity Updated - 8th Feb 2011
@FADS_z
Take a look at the new Motorola Atrix (the rage at CES, coming March 6 on AT&T - an iPhone killer). It has a built in fingerprint scanner, and I can tell you that every major manufacturer of devices is looking at adding fingerprint scanners for simplified locking and unlocking the phone. Add the right web-authentication software to it, and you will be able to swipe your finger on your phone or your PC to be able to authenticate to web sites.

My company, BIO-key, has already begun porting our WEB-key secure web authentication platform client to the Atrix, and when complete in 60 days or so, you will be able to swipe to authenticate to one or more of the major online authentication service providers that are integrating this platform, There WILL be trusted, non-goverment players offering fingerprint authentication in the cloud, and the explosion of mobile devices will be the catalyst for uptake. Standing there pecking in strong passwords while your friends swipe to instantly authenticate in context for secure access to apps, mobile payments, DEA approved ePrescibing of controlled substances (yep, to their credit, the DEA specifically went back and revised their rule last summer to allow ePrescibers to choose a biometric subsystem like WEB-key to secure that process) and BIO-key is already integrated into Allscripts and Eclipsys, Sentillion, EPIC, McKesson, plus most of the commercial enterprise authentication platforms (IBM TAM ESSO, HID/ActivIdentity, Oracle OAM & Passlogix, CA eTrust SSO and Evidian, to name a few). You can also upgrade your laptop's "free" (ie bad) software algorithm to replace it with a better one.

Unfortunately, based on the misperceptions that are expressed in the comments, most people don't realize how big a difference there can be between old or inferior fingerprint systems they may have experiences with, and the state of the art in secure fingerprint authentication today. Tablets and smart phones have had false starts as well, if you recall the failure of the units 5-10 years ago to widely catch on. Did that mean the concept wasn't worthy? No, it meant that the implementation wasn't worthy. Fingerprint biometrics is the same way. Please keep an open mind to how this technology can help make sure that you are the only one who can access your privileges, and make it easier for you, as well. Everyone has the right to a secure identity, and that's what this industry is trying to help achieve.
0 Votes
+ -
RE: Google, Facebook: End Passwords, Get Biometrics. Now!
SecurityThroughObscurity 8th Feb 2011
@FADS_z
A biometric doesn't have to be the only way into your secure accounts - just the most easy + secure way. You can still allow "plan B" access methods, such as smart questions, password plus SMS, etc, but use them only on the occasions that you don't have your fingerprint scanner available. The nice thing about fingerprint authentication is it's the rare case of the strongest authentication ("Who you are") being the easiest.

I see several people raise the question of replay attacks and that you can't create more fingerprints for you if some are compromised. I addressed this below in another reply, but it's important enough simply say here that quality web fingeprint authentication platforms mitigate this threat by creating a secure tunnel protocol all the way to the scanners (even the cheap ones in phones and laptops offer this, and liveness detection to prevent the negative mythbuster exposure that some inferior scanners fell victim to).

As for your fingerprints being of limited supply and vulnerable if compromised, this is a misconception about biometrics, which are different than passwords, which obviously must be kept secret. Intrinsic to biometrics is the idea that you don't have to keep the thing being measured - you - secret. You can show it to the world, but you are the only one who can meet the measuring standard at authentication time.

The misperception is that the fingerprint is the credential, when actually, your finger is the credential. The system's job is to make sure that a real finger is on a real scanner when an authentication takes place. The fingerprint is just an artifact of your finger being scanned, and a quality web fingerprint authentication system will secure that pipeline so an imposter with a perfect image of your fingerprint cannot inject it into the system and claim to be you. The good thing about an all-software platform that is interoperable across all readers is that you can leverage today's enrollment with tomorrow's new scanners and devices that contain them, versus having to start over and re-enroll.

Our customers who enrolled years ago can start identifying their users, customers, and patients over the web using the Motorola Atrix phone on Android OS with our upcoming WEB-key client for Android. Stay tuned!
Probably 60-75% of the time I access Facebook/LinkedIn/Gmail/etc. from a mobile device that doesn't have the capability to do any type of two-factor authentication, let alone biometrics. I would also love to see the false-positive and false-negative rates on the various types of biometric ID, and to what extent that is based on the hardware and software combination used.

In short, I don't think we're anywhere near being ready to support biometric ID across the board, but I'll grant you, we're closer than we were two years ago.
@Real World

Biometrics, unless you are talking military-grade, is VERY unreliable. Believe me, I tested a fingerprint scanner one time that my cousin brought home (they were thinking of using them on the police computers) and we couldn't get the thing to recognize the fingerprint scan as being authorized more than 1 out of 10 times.

This was a consumer level device, but that is the point: it's what most people would use.
0 Votes
+ -
@Lerianis10 We ran a test a while back at my office. The print readers worked well for some people but not for others. I was one of the people it just wouldn't work for (with results similar to your 1 in 10 example) while for others it worked most of the time.

The other thing I have against this idea is that if it becomes widespread, every friend and relative I have will be calling me for their free tech support on it.
@Lerianis10 I agree with you. These dives need to be military grade. The cheapy consumer stuff is not reliable
0 Votes
+ -
Problem has to be the end user and/or linux. I'm guessing both. First it was your facebook that got hacked, now your email. I have to question what exactly you are doing and how you are doing it to be so vulnerable. Biometrics isn't the solution here, you will need to ditch the insecure linux or run it virtualized just like you did with Windows when you thought it was the source of all your troubles. Its obvious linux and you do not match. Time to make that switch to a more secure and reliable BSD box.
0 Votes
+ -
Security
Norm76 4th Feb 2011
@Loverock Davidson I have a CD that I burned from an .iso file I downloaded from the web years ago. It has allowed me to blank the password of *every* single Windows OS computer I have ever tried it on (it is a Linux command line based disk). If I lose the password to one of the Linux computers then I'm out of luck.
0 Votes
+ -
Contributr
@Norm76 Horse poop. With a Linux-based boot cd, you just change the /etc/password file and you're in, provided the disks aren't encrypted.
@jperlow "Horse poop. With a Linux-based boot cd, you just change the /etc/password file and you're in." I'm fairly new to working day-to-day with Linux, so I did not know that. Thank you for the info, I will make a note of it. Though, next time, I would appreciate if you would be more polite in pointing out my mistake (and I was out of luck and had to re-image the machine). It would seem, then, that both operating systems are easy to break into if you have the proper access knowledge and/or tools. If this is the case why do you use Linux to host a supposedly more insecure Windows environment? I'm honestly curious.
@Norm76
There are lots of ways to recover a lost Linux password. If you have physical access to the machine, it's game over, no matter what the OS.
0 Votes
+ -
@Norm76 Actually Perlow is wrong about the file name... On most modern linux systems it's liable to be /etc/shadow, but in some cases is called /etc/master.passwd and various other things. But he has the right idea.

There are things you can do to prevent people from doing that to you, though, like encrypting your startup drive and using a utility like tripwire to monitor for modifications.
@Norm76 ...and plenty of time, its very difficult to keep them out regardless of the OS you are using unless using full disk encryption. And full disk encryption can be done on Windows too (and is on all my systems).
@Loverock Davidson
What not your most secure OpenBSD.
@Loverock Davidson
Agreed. The last time it happened it was all Windows fault. Now that it happened while using Linux as his "main" OS, he says it could have been anything. I'm surprised he didn't blame it on the virtualized Windows session that he hadn't booted yet.

Bert
What if I cut off your finger to use your fingerprint? Or gouge out one of your eyes for a retinal scan?

I think everyone who uses Google Mail should go to Google in person and prove their identity by answering a lot of questions along the lines of, "Which team won the second game of the 2004 World Series?" and "What was the exhaust valve clearance in the first car you owned?"

Plus a short tap dance to the tune of "Finnegan, That's Me."

Otherwise, no access.
0 Votes
+ -
@robin@...
What if I cut off your finger to use your fingerprint? Or gouge out one of your eyes for a retinal scan?

Its fun to make these kinds of comments but the reality is different. I had those same questions but nowadays the biometric actually checks for blood pressure, pulse, and temperature so its not as easy as just lopping off a finger.
@Loverock Davidson

I believe it was Mythbusters that showed how easy fingerprint scanners were fooled.

And I can imagine a glove overlay that will let you etch a fingerprint yet retain the capacitance, transfer the heat, etc.

Not to mention I don't want my fingerprint, retina, etc stored somewhere else and databased....
0 Votes
+ -
Yes but...
cornpie 5th Feb 2011
@Loverock Davidson ...not with inexpensive devices people would be likely to have at home.
0 Votes
+ -
I saw that episode too...
rock06r 5th Feb 2011
@ITGuy8...but I disagree with your statement. Right now, it's EASY to brute force attack accounts for many web services. Heck, there are probably hundreds of scripts to *help* a hacker at this point. But to hack off a finger?? First you'll have to find the user, overwhelm them, and physically attack them. All the while our wonderful security measures (call them "the police", if you will) will be looking for someone running around with a rap sheet of physical violence and a collection of fingers.... ? Not so easy to do from Kamchaka, if you ask me. Outside of Hollywood movie fiction, you can probably count the number of times that this has been successfully done with five fingers (pardon the pun).
@robin@... you can scan all your ten fingers. About the Mythbusters comment below - one has to have an access to your finger and make a copy of your finger print on some sort of a matrix before the scanner can be fooled. SO NO, fingerprint scanner cannot be **easily** fooled. It that would be the case, it would not be used to ID people.
@robin@...

This is security for the AVERAGE computer user, NOT the CIA. Most password hacks of the average user are just situations of opportunity and involve an attacker who isn't even in the same geographical area as the victim of the attack. This, lopping off fingers and stealing eyeballs certainly wouldn't be worth the trouble for this kind of scenario.
This is terrible idea. The problem with biometric is that they cannot be changed. If a site that uses biometrics is hacked, then I cannot use that biometric any where else. What then? It has been proven that you can reverse engineer the seed from the hashed value.

Also: they are not secret! My fingerprints are every where I go. How long before cameras are strong enough to grab enough data from an eye?

Biometrics are tempting because people see them in the movies, but except for physical access via a non-networked scanner for a room with a man-trap, they are a red herring.
@nowen@... YES, I am sure somebody is following you and taking close up shots of your retina... Get real. We are talking about a regular person's access to a web-site with a password or with a fingerprint. What happened here is much, much simpler. A Google employee fired last week decided to retaliate or Google made a mistake a for a while was transmitting all the web traffic as plain text for long enough for somebody to mention what happens. My mother told me that she had to setup a new google account last week because she could not login to her old account. She says that suddenly her Google password stopped working. She does not have any phones, players, or usb drives that she regularly connects. She reads the same web-sites all the time. So it is quite clear that either it was one of the scenarios above or a new virus is in the wild since so many people are affected at the same time.
0 Votes
+ -
RE: Google, Facebook: End Passwords, Get Biometrics. Now!
SecurityThroughObscurity Updated - 8th Feb 2011
@nowen@...
Your comment reflects common misconceptions about modern biometrics - that the enrollment data has to be kept secret, like a password, that possessing a person's fingerprint (whether stolen from a file or lifted from a glass) means that the holder can become you for authentication purposes, your fingerprint data is the credential, making its disclosure fatal, since you can't create new fingers.

All these are understandable, but incorrect assumptions, which is why you do not hear about compromises of quality fingerprint biometrics systems in the consumer markets where they are used.

First, your fingerprint is not the credential - your finger is. No matter how many people have your fingerprint, only you possess your finger, which a quality reader (available even in the ones built into laptops) ensures is really on the reader when read. Unlike the Mythbusters episode which preyed on convenience-over-security solutions like Digital Persona and embedded readers in door locks, real readers have liveness detection and ensure a cryptographic handshake with the server software performing the matching to make it highly spoof resistant. Quality commercial fingerprint platforms (which don't have to be expensive) assume that every bad guy has every authorized enrollee's fingeprints, and then make it impossible to introduce that fingerprint into the secure pipeline from scanner to server. The beauty of biometrics is that you can exposethe thing being measured (you, or a part of you), without fear that seeing the person's fingeprints equals leads to compromise. Thus, just as you need not keep your face a secret to use photo ID as an identification technique, fingeprint ID offers the ability to use your same finger to access many sites, secure in knowing that none of them could use your print to spoof you into another.

Now, my statements sound like a lot of fluff, except my company has biometrics deployed around the world in commercial applications for security and convenience, and I'm happy to provide references at companies who have real-world experience with this technology for years and millions of consumer facing transactions, without experiencing the security breaches that would occur if what the naysayers preedicted is true.

All I ask is that we don't paint the quality providers of fingerprint technology with the same brush as the weak sisters that Myth Busters exposed as worthless. Not all fingerprint products are really enterprise class. My company, BIO-key, is.
biometrics over the internet is an incredibly bad idea.

the security of biometrics is predicated on the party your authenticating with having trust in the entire system, from their own servers all the way out to the biometric sample acquisition hardware. by definition they cannot have that trust when they don't have complete control over all parts of that system.

the only secure biometric installation is one where the authenticating party has no access to the system except the biometric reader, and then only if there's a guard present to make sure the authenticating party doesn't try anything funny with the reader.
0 Votes
+ -
RE: Google, Facebook: End Passwords, Get Biometrics. Now!
SecurityThroughObscurity 7th Feb 2011
@imaguid
Please see my reply to the post above. I've been in this industry selling biometric solutions into successful, secure commercial applications running at some of the largest, most security oriented companies in the world, and we do exactly what you say you can't do - secure the entire pipe with 1024 bit elliptic curve encryption. Even the path from the scanner firmware to the software in the client has a cryptographic noonce handshake.

We assume that the user's USB is being sniffed, their browser is hacked, their client PC is compromised, their network is being sniffed, their app server and server application is compromosed, and we protect the data from intrusion or injection of data at any point along the way.
Things to note:

-Biometrics, if not done right, are suspect to replay attacks. Once your biological info, no matter what the original form, is 1s and 0s, than you have to be wary of people copying those 1s and 0s.

-There are . . . gruesome ways to get some bio signatures.

-There's gonna be the case of people w/out certain limbs for some reason or another (birth defect, machinery accident, war injury, etc) - need to account for them.

"Fingerprint scanners are inexpensive, ranging from $40-$50 retail if you want to add one to your PC."

"inexpensive" is subjective. For some people, it's not a big deal, for others, it can buy a good amount of food.

"Built-In cameras in laptops and smartphones with high-resolution CCDs and constantly improving macro capability and miniaturized optics also could make retina scan on portable devices and PCs an affordable reality within a number of years."

Eh, retina scan is more than just a camera, if I remember correctly. You have to light up the inside of the eye to make the retina visible.

Besides, while resolution is certainly increasing, I don't think it's increasing *that* fast.

"These solutions could also be combined with RFID implants and/or voice print identification to have multiple points of identification, in order to minimize the risk of access due to biometric forgery or coercion under duress (such as being forced to authenticate under gunpoint)"

I don't know about you, but unless I'm holding state secrets - I'd much rather let the person have the info than have a bullet in my brain.

And no thank you to an RFID chip under my own skin. I'd rather have it in some jewelry or in my wallet or in my phone or somewhere else besides under my skin.

" What we really need is a standardized API"

That's really the problem that needs solving, and biometrics won't necessarily solve it (what if there are multiple competing APIs for the biometrics stuff??).
0 Votes
+ -
RE: Google, Facebook: End Passwords, Get Biometrics. Now!
SecurityThroughObscurity 7th Feb 2011
@CobraA1

There are competiting, proprietary APIs among fingerprint biometrics vendors. There are also industry standard APIs, such as BioAPI, which normalizes the interface across all products. That makes your developers happy.

But what about all the proprietary enrollments you create, which lock you into a specific reader? The real asset in any systme is the vetted enrollments, not the APIs, so choose an software platform that provides you with a highly accurate, but device interoperable algorithm that lets you move from reader to reader with complete recognition of you on any of them.

These problems are already addressed, but it is obvious that my company, BIO-key, is not doing enough to herald our sucesses and make sure that people find us.
0 Votes
+ -
IPSec in IPv6 will fix a lot of this. Once all of your devices have their own unique, public routable IPv6 IP address, all you'll have to do is tell your Facebook and Google accounts what IPv6 addresses are allowed to access your Facebook and Google accounts. Problem solved.

So in a few more years, we wont' even need biometrics....or a password for trusted devices.
@VRSpock Not exactly.

First of all, NAT has benefits well beyond opening up more address space, so I don't think it's going away.

Second, if we want IPv6 identifiers to be used for logins, we must absolutely make sure they can't be spoofed and are visible even behind a device using NAT.

Last, what happens if a device gets a new owner, or is stolen? What happens if a user wants to use a new device? There has to be mechanisms in place for transferring, cancelling, and creating profiles associated with the IP addresses.

IPv6 may make it easier to create a solution, but is not by itself an actual solution.
0 Votes
+ -
@VRSpock Not really... I share my desktop with my girlfriend pretty regularly, and when I'm on the go I can't guarantee that I'm not going to sign in from a netcafe or a random mobile hotspot.
0 Votes
+ -
Great idea!
crazydanr@... Updated - 4th Feb 2011
Let's give fb and google my biometric information. Since they would have to store the biometric data to authenticate against it, I would love that information on a server up in the cloud.

Both companies can be trusted to use this data only in an ethical manner that maintains my privacy and prevents intrusive marketing and advertising.
0 Votes
+ -
Your biometric data is already public if you don't wear a burka
SecurityThroughObscurity Updated - 14th Feb 2011
@crazydanr@...
Everyone needs to realize that biometric data is not like passwords which need to be kept secret. You and your biometric data are already public, and biometric authentication systms assume that to be the case. Putting a picture of you in a database doesn't introduce new risk that wasn't already there by you walking down the street in public. The same goes for your fingerprint images, even though you might think they are harder to collect. You leave them on everything you touch, and as you point out, you can use them to prove you are you to various counterparties, so they can be put into a database. All that database allows the holder to do is to tell if you are still you when you come back. It does NOT mean that they can take your enrolled fingerprint data - whether image or template - and submit it to another biometrically enabled site and say, "I'm him." The biometric system's job is to make that impossible. Crappy products don't do that, and Mythbusters makes the naysayers say the whole industry is similarly vulnerable. Not every biometric system is the same, not all are vulnerable in this way.
There are a lot of different ideas out there for user authentication. The problem is, eventually they will all be cracked one way or another. Some of them through Social Engineering, some of them through more powerful and smarter processors. @Jperlow wants Biometrics, as @itguy pointed out, Mythbusters was able to get around them. For the most part, that is true. It depends on how much money you want to throw at it and your acceptable risk. Of course the cheaper scanners will be easier to get by, depending on how many points of reference they use, resolution of the scanner, and even the software/firmware being used. (I?m not 100% on this as it is not my field, but sounds right).

CobraA1 mentioned sub-dermal RFID tags. An interesting idea, but probably also vulnerable to some sort of spoofing. RFID would have to transmit via low powered radio waves. While I?m not exactly sure how these work, I?ve read a little about them. They would probably have a unique number or serial ID coded to them similar to a cellular ESID. Ah, but ESID?s can be cloned; as I?m sure these could be also.

Going a little aside here, I saw a video by a group that does Tiger Team security assessments. Basically, they go in and see how hard it would be to break into you system kind of like pen-testers. They were contracted by a high profile jewelry firm to see if they could break in without being caught. The firm used high tech proximity badges to enter the back rooms of the store. The Tiger Team was able to walk by the CEO on the street outside the shop and ?accidently? bump into him, and clone the proximity badge in his pocket. Using that and an easy to guess pin code because of worn buttons, they were able to gain full access to the store. I can see something similar with RFID chips and pinpoint directional radio beam, like some hackers have been able to do with bluesnarfing.

This brings up two factor authentication. This is called ?something you know? (UserName & Password) and? something you have? (OneTime Pin Code or Smart Card or other token) Next up, some companies use RSA SecureID style keyfobs. While not easy, these have been cracked. Many banks use have started using One Time codes that they will SMS to a user?s cell phone. There is now a smartphone botnet that will catch these and retransmit them without the user?s knowledge. User thinks they never received the SMS, hacker uses it to log into bank site before the time is up. DOD uses smart cards, internally, Banks use USB tokens with signed certificates, I?m sure I?ve missed something.
A lot of places us security questions in case you need to reset your password. Well, how many of those answers are on your Facebook profile. Mother?s Maiden Name? Check! Maybe you didn?t put it there, but she did, so her old highschool friends can find her?and then you have relationship link to her. Date of Birth? Quite a few are out there. Favorite food? Well, you tweeted 3 times last week that you had Domino?s. I sat through a talk at ShmooCon (A Hacker conference in DC last week) where the presenter showed how easy it was to get information out of a eVite by using user & event ids in the URL. One of the things I noticed was all of a users profile information would show up, including the answers to the password reset questions. And eVite seems to have extra questions, probably to help with a profile when planning a party. I?m sure eVite is not the only site with lax security.

I had a few other ideas, but I think I?m starting to ramble here, so I'm going to end it here. I think this is enough for discussion.
0 Votes
+ -
Are you INSANE?
wolf_z 4th Feb 2011
How the heck can you claim to be a tech person and not know why biometrics is a horrible authentication technique?

First, the bioscanner is not going to teleport your finger to Google, it's going to send a numeric representation of your print.

Over the internet.

Yeah, what could possibly go wrong with that?

Not to mention once someone has that magic number you're screwed, blued, and tattooed--your identity is theirs and you are FOREVER hosed. You can't change your fingerprint, right?

Biometrics. What a stupid idea...
0 Votes
+ -
RE: Google, Facebook: End Passwords, Get Biometrics. Now!
SecurityThroughObscurity Updated - 11th Feb 2011
@wolf_z Your concern is valid, but represents a common misconception about biometrics that I address below to SolidWater's similar comment: why the "you can't change your biometric characteristics" is misapplied in this context.
Unlike a passwords which must be kept secret, biometric authentication presumes that the thing being measured (your face or finger) is exposed to public view and thus cannot be kept secret. The systems which do the authenticating need to have good liveness detection to know that they are not being spoofed with a copy of the person's characteristic, along with a secure pipeline from the capture to the server to prevent a bad guy with a perfect copy of your biometric data ( from being able to slip it into that pipeline and say, "I'm him." Quality biometric systems do that - in our case - with 1024 bit Elliptic Curve encryption, with a key that changes on every transmission.

The problem is, there are a lot of terrible biometric products out in the market, which give people the impression that the technology in general doesn't work or is insecure. Mythbusters spoofed a terrible subset of products in the market, and now, people understandably point at that as the reason not to trust biometric solutions.

There are many analogies in technology history to look to - early cell phones could be cloned. Do you worry about that now? Would it be fair to say, "You should not use a cell phone - ever - because I once saw that they cloned one? Not when the technology has advanced to the point where it has, where our software system presumes that there is a hacker with your fingerprint in perfect digital form, and has a sniffer on the USB, has compromised the browser, the network, the app server and even the application, but will not be able to spoof our biometric authentication process.
Yeah, let's give Facebook and Google access to our fingerprints. That sounds like a great idea doesn't it? What exactly happens when someone steals an image of your fingerprint again? Is facebook's security really that good? No.

This may have been the dumbest security suggestion I have ever read, and I've read posts on the internet suggesting that Windows ME is a good network operating system.
0 Votes
+ -
Biometrics don't work
jshaw4343 4th Feb 2011
I've seen several clients put in biometric systems, only to rip them out a couple months later. They just don't work that well. Or there are a lot of cheap products that can easily be fooled or by-passed. And the ones that do work, are not very efficient. They tend to take a long time to read and authenticate which can be frustrating for users.
0 Votes
+ -
@jshaw4343
I absolutely agree that there are terrible cheap (and terrible expensive), products in the biometrics market. It's a hype-filled space, and you really need to look for the companies and products which are not a "me too" solution, trying to attract investor money or puttying over the holes in their security. I think that my company is one of the ones that delivers, doesn't get ripped out, and bucks the claim you make that this doesn't work.

Why do people conclude that "Biometrics doesn't work"? Because the accuracy of most algorithms that are in the market is around 1:1000 to 1:10,000 False Match Rate (FMR), which means that they make a "someone else is you" mistake 1 out of every 1000 to 10,000 match attempts. The result is that the threshold has to be bumped up in higher security applications, meaning that the False Reject Rate ("You are not you") goes up, making people think the software can't recognize them.

State of the art software in the fingerprint arena has been tested and certified by NIST for accuracy. Our product has accuracy that is orders of magnatude higher than the average product in the market, and has been tested by NIST to back that claim up. The algorithms you have had bad experiences with (such as the one that comes stock with laptops or is in the POS terminals), is not tested by NIST. Beware of products that brag about "winning international competitions," because now that NIST offers free testing using millions of prints, no serious fingerprint company participates in any other test or competition, so "winning" means beating up on weak players who don't submit to NIST because NIST will publish their results for all to see. Like I've said, a very hype-filled industry.

We make software that is independent of fingerprint readers (but knows the security characteristics of the various readers we support, allowing restrictions to keep vulnerable readers out of the system). That means that as reader technology advances, our customers can move to the latest and greatest without having to re-enroll everyone. We can positively ID any person out of a database of 100s of thousands or millions in just a few seconds, on standard PC server hardware that is not very expensive at all. Readers don't have to be expensive to perform very well and be secure. We have customers who pay less than $1 per enrolled person to be able to positively ID them at any time in the future - forever. Is that expensive compared to mailing a SecurID token or smartcard to everyone?
0 Votes
+ -
NO!
Solid Water 4th Feb 2011
I want my privacy, I do not want somebody stealing my biometrics, filling in BIOMETRICS database with stolen information and then selling this information to third parties. I think that there is enough information already collected on the Internet about each and every person.

How long would you think it will take the bad people in this world to start implementation?

Honestly, being in computing since 1980 I want privacy on the Internet.

I do not like the fact that Facebook sent e-mails to everyone on my mailing list inviting to become my friends when my wife created Facebook account from my Yahoo e-mail.

This is the cry of my soul. wink

YMMV wink
0 Votes
+ -
RE: Google, Facebook: End Passwords, Get Biometrics. Now!
SecurityThroughObscurity 11th Feb 2011
@Solid Water
Your comment is empassioned, and I'm glad to tell you that biometrics does not work the way you contend. If biometric enrollment data were like passwords, in that knowing or possessing it gives you access to any other system which relies on the same biometric, then the technology would be useless.

Unlike a passwords which must be kept secret, biometric authentication presumes that the thing being measured (your face or finger) is exposed to public view and *not* kept secret. The systems which do the authenticating need to have good liveness detection to know that they are not being spoofed with a copy of the person's characteristic, along with a secure pipeline from the capture to the server to prevent a bad guy with a perfect copy of your biometric data (your scenario 1 - someone who hacked a server and has a copy of your enrollment from some other system, or 3 - where someone lifts your prints from you, a glass, etc) from being able to slip it into that pipeline and say, "I'm him." Quality biometric systems do that. the problem is, there are a lot of terrible biometric products out in the market, which give people the impression that the technology in general doesn't work or is insecure. Mythbusters spoofed a terrible subset of products in the market, and now, people understandably point at that as the reason not to trust biometric solutions.

There are many analogies in technology history to look to - early cell phones could be cloned. Do you worry about that now? Would it be fair to say, "You should not use a cell phone - ever - because I once saw that they cloned one? Not when the technology has advanced to the point where it has, where our software system presumes that there is a hacker with your fingerprint in perfect digital form, and has a sniffer on the USB, has compromised the browser, the network, the app server and even the application, but will not be able to spoof our biometric authentication process.
You really trust google and facebook to look after your fingerprint data???? What have you been smoking????
0 Votes
+ -
Contributr
@andrewk7 They won't need the actual fingerprint data.
Or you could insert your "you know what" into a secure device attached to your computer and have your size measured. Incorrect size means no access! Take this as another vote against the use of biometrics!
What follows is a completely "pie-in-the-sky" .. "thinking outside the box" idea for home PC security. (It also means I should have retired to a restful sleep long before I submitted this post. Grin)

When it comes to high security access, Hollywood always shows us a "two way" access solution. For instance, the Government sends out the "war code" to its military. The military has to confirm that code before it has access for a follow-up action.

In like manner, you try to access G-Mail. Google sends you out a code. You have to supply the correct counter code before you gain access. Obviously, for the security protocols to remain as robust as possible, this code pair changes every time one tries to access G-Mail. (BTW, I use G-Mail only as an example. If one wished to access his online banking info, a similar but different dual code access would be required.)

Don't ask me how this would work for individual PC systems. I haven't a clue. But it seems that this type of dual code security access has worked for government programs over decades. Could a similar approach be modified to work for the Jason Perlows of this world?
0 Votes
+ -
Contributr
@kenosha7777 Well, the other solution is giving everyone RSA cards like many financial organizations do for their most important apps. But you can still lose an RSA card, and getting one replaced is very, very inconvenient. My guess is you'd have to wear it on your person, like a digital watch. You can still be forced to use it under duress, but then again, you can be forced to log in under duress now anyway.
@jperlow You could theoretically be forced to do practically anything under duress, including sticking your finger on a reader. Biometrics aren't going to change that.

And my info isn't worth so much that I'd take a shot in the head for it, so a biometric that attempts to detect my duress would likely only hasten my death, so no thanks.

I'm personally all for giving people ID that they aren't forced to carry everywhere. IMO forcing people to be chipped or IDed regardless of their wishes is the first step away from freedom, and likely won't go well with some religious people who believe in a "mark" in the last days.

I should have the option of not having my ID everywhere I go, thanks.
Good day to confirm this comment I would appreciate T h e b e s t o f Z D N e t d e l i v e r e d your website very nice to everyone Yes, Oracle is the only one with shared-disk architecture, but that is there advantage. It means you can add or remove nodes and the database lives on. In a shared nothing architecture, if you lose a node, you lose the system. I'm sure Oracle appreciates EMC highlighting their advantage.I also desire to signal in your RSS feeds. Thank you as soon as once again and maintain up the great operate Awesome post! Thank you very much || thanks for nice content this is really benefit to me.

Join the conversation!

Formatting +
BB Codes - Note: HTML is not supported in forums
  • [b] Bold [/b]
  • [i] Italic [/i]
  • [u] Underline [/u]
  • [s] Strikethrough [/s]
  • [q] "Quote" [/q]
  • [ol][*] 1. Ordered List [/ol]
  • [ul][*] · Unordered List [/ul]
  • [pre] Preformat [/pre]
  • [quote] "Blockquote" [/quote]
ie8 fix

The best of ZDNet, delivered

ZDNet Newsletters

Get the best of ZDNet delivered straight to your inbox

Facebook Activity

White Papers, Webcasts, & Resources
ie8 fix