Microsoft Defends Passport To Privacy Group
securitas writes: "CNET reports that Microsoft is defending Passport as safe and secure in a presentation to the Center for Democracy and Technology. Other organizations such as the Electronic Privacy Information Center, Junkbusters and even the U.S. government may be lobbied by MS this week to fend off a Federal Trade Commission complaint filed by 15 consumer and privacy groups that charges unfair and deceptive practices."
This says it all:
"One of Passport's greatest security weaknesses may be the single sign-on process, analysts said. The single point of entry could also be a single point of failure. Since the ID is always an e-mail address, someone looking to break into an account might easily obtain half the information needed to do so."
Because people usually don't pick very secure passwords, it's better to have multiple passwords so that an evesdropper or other malicious person can't crack into all yur accounts. U of I just made people intentionally set all their 3 or 4 passwords instead of just giving them one the applied to all 4 (although most people tend to choose the same password for all their online services anyway)
Also, because Passport's trying to incorporate a lot of information in one place that used to be distrubuted in many different places, if some one hacks into Passport, there goes all your privacy.
F-bacher
James Tiberius Kirk: "Spock, the women on your planet are logical. No other planet in the galaxy can make that claim."
Passport is definitely an easier solution for consumers than any alternative yet presented. Having all your information stored in one central location is definitely better than having all your information stored all over the place. Microsoft also has a lot more motivation and resources to protect it than Joe Random Vendor.
The problem is that they haven't had any success protecting it anyway. To be completely fair, neither has anyone else. The other difficulty is that although I would trust MS rather than JRV to protect my data, the necessity of distribution and interaction opens up a whole new class of security holes that no one has even thought of before.
The unfortunate truth is that right now the only way to protect your privacy online is not to give out any information, and that Passport will do exactly nothing to remedy this situation.
Even Slashdot wants to hide some things
So these privacy groups get worried about Microsoft's Passport leaking information when the biggest leaks of personal info are from fallen dotcoms and stupid e-commerce web sites? People, when you are paranoid, at least be paranoid to everybody, not just to Microsoft.
¦ ©® ±
Just last month, Microsoft changed the service agreement for their passport system to require only an email address and password to sign up. Did Microsoft do this without any armtwisting? No. Did they do it, though? Yes.
Just keep the pressure on them up. They're going to go ahead with some sort of service no matter what, but the amount of opposition they face now will determine how many of these concessions will be made "voluntarily". That way, even if the FTC doesn't come down with a favorable ruling, we won't be completely left out in the cold.
Incidentally, msnbc also has some coverage. A disinterested and impartial news source if there ever were one... or not, as it were.
For those that are interested here are links to the:
Passport EULA
Passport Privacy Policy
Privacy advocate: "So, you are trying to set yourself up as the one definitive source for our personal information online. Let's talk about your record: Hotmail backdoors, Code Red, Melissa, IIS, and Kournikova, among others, are horrible things which have been influenced by your poor implementations of products. And you want to have even more power?"
Microsoft PR guy: "Try to think of those as valuable lessons we have learned to make Passport more secure...
Information leaking from one site is annoying, esp. if it's something like a credit card number, but it's nothing compared to aggregated information being leaked.
As a silly example, let's say you buy rat poison. No big thing, people buy it all the time.
Let's say you buy a book about "perfect murders... and how they were caught." No big deal, people buy true crime books all the time.
Now let's say you recently bought a bunch of lingerie. And had it delivered. But not to your home address. You're having an affair, sleazy, but not unheard of.
Now finally let's toss in the fact that you just consulted a lawyer. A divorce lawyer. One who specializes in breaking prenuptial agreements.
Suddenly things are much more interesting.
Most of us aren't planning to murder our spouse, or even to look like we're thinking about it. But it's certainly possible for mindless data aggregation to cause people to jump to the wrong conclusion. E.g., you bought a couple books on alcoholism, and a few cases of wine? You obviously have a problem, don't you. (Nope, the wnie is a gift to newlyweds and the book is to help me understand if my nephew needs help.) Etc and so forth.
Even with all of this information centralized with Microsoft (and make no mistake that the Passport/Hailstorm system will not collect this information), my biggest concern isn't that it will be leaked. My concern is that it will have bogus information feed into it. There's a nice market opportunity for nasty companies to put bad information into these records, then offer to clean it up for you. For a modest price, of course. All of the potential damage of a credit report, but with none of the legal safeguards.
Of course, that same problem exists today with the aggregated data provided by from credit card companies, but again it isn't a *single* point of failure. Even if you crack Citibank (still the largest CC issuer?), it does nothing about the hundreds of millions of people who don't have Citibank cards. But crack Hailstorm and you'll have information on almost everyone online.
For every complex problem there is an answer that is clear, simple, and wrong. -- H L Mencken
this is such a classic microsoft-ism: thinking up a really good idea, and totally fucking up the implementation ([d]com, ole, activex, etc).
.NET has shackled it to possibly the worst authentication possible.
.NET?
.NET platform to the username/password "security system" is about as intelligent as locking your car with duct tape, and will probably be about as effective.
what I can't figure out is why this company, which is supposedly on the brink of launching this massive, multi-tiered platform that is
I mean, come on, the username/password combo was maybe reasonable in the days when everyone had exactly one shell account. but today when everyone is expected to remember a user/pass combo for every one of a dozen or so websites they want to log into, the weakness of this paradigm has hit pretty hard. simply put: people can't remember them all, which means they either write them down lots of places (prett damn insecure) or use the same username/password for each account (even worse).
and MS has made THIS the lynchpin of their security model?
why couldn't MS use some of their much vaunted "monopoly power" to "leverage" an authentication system that actually matched the sophistication of the rest of
my suggestion: the medium which most people are accustomed to carrying that is intimately tied to their financial and personal data is the credit card. my MS "Passport" could be a physical smartcard that held authentication data, encryption keys...hell, anything. each copy of XP (and each bundled OEM copy) would include a small USB device that could read this card, maybe that was designed to mount onto the side of the monitor so it would stay out of the way.
YES this would be a major move, and it would stir things up a little. but when it is clearly called for, WHY NOT? people would just carry another little card in their wallet, the reader device would be small and dirt cheap (in that volume, most anything is) and in a year we would forget what we did without them. we have calling cards, and credit cards,and ATM cards...where is my computer card?
in any case, tying their much-heralded
When you sign-in to Passport there are two checkboxes...
One says 'Sign me on Automatically'. If you check this, a cookie is stored that remembers to authenticate you from then on.
If you don't check this box(which is the default condition), then a cookie is created and stored which remembers your username. But the authentication information is stored as a session cookie which disappears when you close the browser.
There is a second checkbox. It says 'I'm using a public computer'. This stores a session cookie on your machine for both the username and authentication.
Once you have closed the browser, the session cookie is gone and you no longer authenticate automatically, nor is your username auto entered for you.
So while I understand your concern, Microsoft has provided two checkboxes which alleviate this concern. Neither checkbox is on by default which means the default behavior is to remember your username only.
If you have a better solution to this problem, I'm sure we'd all appreciate hearing about it.
BTW, the paper you linked to has much better explanations of problems Passport might have then what you wrote about. Man in the middle type attacks that involve redirecting DNS, etc.
Unfortunately, that's just not true. Usability research has shown certain facts about passwords again and again. In particular, as soon as you start forcing users to remember several passwords, they immediately start using obvious and easy to remember passwords, or writing them down in a readily accessible location. Clearly, this does not improve security.
Having a single sign-in, with a single, genuinely cryptic ID and password, is far more secure than twenty different authentication schemes for different facilities. Of course you rely on the keeper of that information to keep your data in a trustworthy fashion, but you have that problem anyway. At least with a single secure sign-in the average five year old can't guess everyone's ludicrously simple password.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.