UK gov admits gamble on massive net snoop plan The Home Office has privately conceded that its plan to store details of every internet communication may not be possible - and that it has pinned the multibillion pound project's hopes on snooping technology not yet developed. Officials working on the Interception Modernisation Programme (IMP) made the admission last week in …
For those of us who actually use DPI devices, we know this is entirely possible with even a moderately powerful device from most manufacturers today, this has been so for a while.
These devices can understand many different protocols and even their content in many cases. Logging the establishment of the connection is trivial to say the least.
Pulling all this information together on a per-user basis is a task for another device, and a little more time consuming and tricky.
Basically these chaps are lying or incompetent.
If you want to monitor someones communications, you get a warrant, the ISP will route their connection to your computer and you will be able to monitor it. The ISPs can route connections however they see fit, it's a configuration on their router. All traffic all ports. This is how they route users who've gone over their bandwidth limit to a block page.
You have never had the power of mass surveillance before, the databases weren't there, the deep packet inspection isn't there now for the volume of traffic concerned. You also can't justify spying on everyone all the time. The technology didn't exist and it is expressly forbidden by their fundamental rights.
Jacqui Smith is gone, her mindless obeyance of anyone in a suit is gone. Now you actually have to talk sense. You can no longer pretend that somehow you can't monitor ONE internet connection without monitoring EVERYONE ALL THE TIME.
When it was pointed out that the anti-terror stop and search power was being used against hundreds of thousands of people, with zero conviction rate for terrorism, the Home Office spokesman said this:
"Stop and search under the Terrorism Act 2000 is an important tool in the on-going fight against terrorism.As part of a structured anti-terrorist strategy, the powers help to deter terrorist activity by creating a hostile environment for would-be terrorists to operate."
The random searches create a hostile environment for ordinary people, they are complaining, and he turned that around and linked the hostile environment to terrorists..... see *terrorist* find it a hostile environment, and *you* find it a hostile environment therefore what does that make you???
But it's just talk, it has no logic behind it. Now that Jacqui has gone, there's a chance for a fresh start, some logic and reasoning to policy, rather than an emotional moral panic.
Whitehall must simply not 'get it'. They're ultimately up against cold, hard mathematics, and they won't win.
They either look at everything (not merely "communications data", but all the contents as well), or they leave criminals (terrorists, etc,) with opportunities to keep their communications beyond surveillance.
Imagine Alice, Bob, Carol and Dave are part of a secret communications network. Alice writes a secret letter to Dave, puts it in an envelope addressed to Dave, and then puts that envelope in another envelope addressed to Bob. Alice posts that letter to Bob. GCHQ log the fact that Alice has sent this letter to Bob.
Bob gets Alice's letter, opens it, and sees it's just a sealed envelope addressed to Dave. He puts a cross on the back of the envelope, and puts it in another envelope addressed to Carol. He then posts it. GCHQ log the fact that Bob has sent a letter to Carol.
Carol gets Bob's letter, and opens it to find an envelope addressed to Dave. There's a cross on the back, which she knows means it's already been forwarded once. She puts the envelope in a new envelope, and addresses it to Dave. She posts it, and GCHQ log this fact.
Dave gets a letter from Carol. He opens it, and finds it's an envelope, addressed to him. He opens it, to find it's a secret letter from Alice.
The secret letter contains a fragment of another piece of correspondence, encrypted by some method. The letter also contains instructions on how to forward that encrypted fragment on towards its final destination, and how to re-encrypt it for that purpose.
How deeply do GCHQ have to inspect these network packets in order to truly keep track of everything? How many layers of envelopes must they open? How much material must they actually record? How much of the contents must be kept on file in order to truly keep up with secret communications?
For IMP and the like to actually work, it can't be just "communications data" that's logged. The actual contents have to be inspected as well, just in case there are hidden layers of protocols within the actual packet payloads. We either completely lose privacy, or the whole exercise is a gigantic waste of money. Anything short of a total loss of communications privacy, and the criminals can still keep their communications beyond inspection - which is where we already are if we don't do this IMP stuff in the first place anyway.
Why spend billions ending up back where we already are? Are we really going to sacrifice all our communications privacy? I haven't even mentioned steganography, yet!
The State needs to take a fundamentally different approach. IMP sounds like an attempt to make this newfangled internet stuff be more like old fashioned telephony and the like. It's like the recording industry wanting to hobble the internet so as to preserve their out-of-date business model. It is, instead, the State that must radically change its approach.
Instead of trying to "maintain" capabilities (and they may well genuinely believe that's all they're trying to do, even though they're clearly doing stuff never done before with snail mail), they need to undergo a revolution themselves. They need to accept that criminal organisations, including terrorists, will continue to have secret communications arrangements. Instead of trying to make that fact go away, the State and its agencies need to find ways to fit into this reality. That means finding ways of using modern technology, the internet, etc, in ways that aren't ultimately futile.
They need to think outside their increasingly out-of-date box.
<...raised the issue of encryption at the meeting last week. Mass take up of scrambling technology - which would render a lot of internet communications data unreadable by DPI probes - had been feared by government before, officials said, but did not materialise.>
That'll be because the global monitoring of every communication was not on the horizon before. I've said it before and I'll say it again: It's none of their fscking business.
Government policy: A digitally connected Britain, to make it easier for the government to track our every thought!
Apropos of nothing: why is not *every* internet connection HTTPS?
Isn't it obvious - there is much more gain in the numbers game for showing how efficient you are - from catching people who would otherwise be law-abiding if you can catch them breaking small rules they know nothing about and would have difficulty even realising they were breaking such (such as 31 mph) then there is on catching real criminals
And that's the rub, isn't it.
It is relatively trivial for anyone who *wants* to avoid Gov't DPI snooping to do so. So this system is going to be next to useless against all but the stupidest criminals and terrorists. This fact is so blatantly obvious that even the most technically illiterate Gov't minister must be aware of it, or at least have been informed of this fact by those that *do* know.
So, given that IMP will be all but ineffectual against criminals and terrorists, why bother?
It can only boil down to this government's insane desire to monitor and control the lives of every British citizen. It is aimed at catching people committing inadvertent misdemeanors, not at stopping organised crime or terrorism.
This is not to catch people breaking the law it is being slowly put in place to police to normal man. For years governments of the world are trying to make us stay indoors, High tax/increasing cost of living. Fear of attack, fear of the people we live around, it is working people are staying in rather than going out. At the very least outside socialising is being dramatically reduced.
But then where do our social interactions move to. Thats right the internet, with the advent of facebook, twitter second life, and other social gateways the governments need to try and control the information flow through them.
The old saying if you have nothing to hide you have nothing to worry about, 'yes you do'. We are being watched when we walk out of our house, how many camers in the UK have seen you today just going to work, shopping etc...
This can all be used to create a social blueprint of the UK/wold populous, it will allow better control as it matures. It all sounds very far fetched but how far off is it really..
I was under the impression that other than the obviously "OK meet up outside the bookies, tomorrow and 10am and then go blowup the local ASDA, don't forget to bring the explosives" the content of the communications wasn't really that important.
http://www.lrb.co.uk/v30/n16/soar01_.html
Now people ought to sit up and realise, that all these ideas that Jackboot Smith came out with which appear to have had the backing of the Unelected Prime Minister is nothing short of mismanaging tax payers money.
This is a blatant example of Ministers "gambling" with tax payers money. Forecasting the costs for a spy technology without having the technology in place though hoping it would one day catch up is yet another example why this current government is not fit to be in office.
Are the latest DPI devices able/allowed to intercept the initial key exchange send when negotiating a secure connection?
If so, then I'm surpised and a tad worried (for "we're not inspecting the content", this seems to be pushing it).
If not, then why can't the malicious people just buy a VPS outside of the country, SSH/VPN into it and move on with their lives?
*Confused*
"ISPA said IMP officials had added that they expected the new laws needed to legitimise the system would take "a number of years" to pass"
So currently it's illegal then and it's forseen that it will be for a number of years? Does that mean our tax money is being spent on something that's currently against the law? Surely that should result in a rather swift visit to jail for whoever approved it?
Less seriously: They'd get a lot of "deep inspection" on the way into jail and they might even be a spot of "packet inspection" in the showers too...
Oh, how quickly everyone forgets.
http://www.theregister.co.uk/2007/09/06/advise_cancelled_data_mining/
The "cancelled" here *WAS* optimistic. What about TIA before that, or Carnivore before that? This technology has been around, and commerically marketed (FFS!) for/by credit companies, banks, insurance companies - all to "determine risks" - Pshaw!!
I am sorry, my brothers in Blighty. America *does* give technology back to HM's subs.... but it just isn't the technology any free, public populace wants.
"have pinned the multibillion pound project's hopes on snooping technology not yet developed"
Hahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahah Breath.... ahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahbreath....
Isn't it becoming rather obvious that all this police state stuff has rather little to do with who the Home Secretary is, or probably even who the government is? Different incumbents may make different noises but whoever the current mouthpiece is, the plan to monitor everyone's every waking moment just keeps rumbling on...
Isn't it clear that the nub of the problem is the Home Office and that changing governments probably won't change the policy either. Until we have someone in power who's prepared to take the HO apart piece by piece and only keep the bits we actually need, I think we can only expect more of the same.
It doesn't matter if they *do* intercept the 'key exchange'. That's the whole point of PKI. I send you my public key, you encrypt a symmetric key with it and send it to me. The only person that *can* decrypt it is me (with my private key). We now both have the same symmetric keys and can send encrypted data to each other.
The above is the general routine used in all PKI based systems, SSL/HTTPS included, though there are additional complexities with cert authorities.
The State wants intelligence because it has no intelligence.
The State does not have the intelligence to obtain the intelligence it wants.
In other words, it has neither the means nor the end.
Given that to fill an array with random bits entails a function with quite a lot of intelligence behind it, that is quite remarkable.
I, like zillions of others use bog-standard gmail which by default requires SSL to connect to a non UK server. Will the govt be able to spy on my mail ? If yes, then that is a privacy outrage. If no, then the whole spying thing is a huge waste of taxpayers money, and again an outrage. Sure, google will probably hand my mail over to the uk govt if said govt request, but I what I actually want to know is, will something as routine as ssl defeat this snooping technology and if it will, then what is the point of it (the snooping technology) ?
"The government aims to extract coherent information about online behaviour from raw internet traffic."
This is basically.
Find a terrorist.
Study his comms pattern (type, numbers, service providers etc)
Match pattern agains all customers.
Locate all other terrorists with this pattern.
Repeat for each additional terrorist you have on file but whose comms have a different pattern.
The National Science Foundation report on this idea (reported int he Reg) reckoned it was a long shot from Mars to work. It should never be tested on live data without extensive testing on dummy data and the number of false positives would be huge.
And I would guess what could be described as an n-dimensional graph matching problem (sort of n-dimensional egrep of a large n-dimensonal data file) could get quite compute bound.
But it's a great way to spy on your population. If your a delusional paranoid who believes that no one likes you. IOW it's not about defending the electorate from terrorists. The electorate *are* the enemy.
Does any one here know anyone like that?
A number of ISP's do it already, Universities are doing it (well they are ISP's in law anyway) in a vain attempt to control their academic's telling the truth about University management, hell pretty soon you won't be able to put anything on the web that might offend someone...So much for freedom, or even 'academic freedom'.
"It doesn't matter if they *do* intercept the 'key exchange'. That's the whole point of PKI."
Are you sure?
"I send you my public key,"
But Malcolm, the Man-in-the-Middle, intercepts it. He generates his own public key, and sends that on as if it's yours.
"you encrypt a symmetric key with it and send it to me."
Which Malcolm intercepts, and decrypts using his private key. He now knows the symmetric key. He encrypts that symmetric key using your public key, and sends it on to you.
"The only person that *can* decrypt it is me (with my private key)."
Which is fine for Malcolm, since he already knows what the symmetric key is.
"We now both have the same symmetric keys and can send encrypted data to each other."
And Malcolm can decrypt all of it, without either of you knowing.
"The above is the general routine used in all PKI based systems, SSL/HTTPS included, though there are additional complexities with cert authorities."
Ah, yes. Handy for guarding against such things as Man-in-the-Middle attacks. But how do you know the certificates aren't also being intercepted and tampered with?...
1. Reg readers produce pretty brochures, website and PowerPoint presentation for the technology.
2. Get government to spend the money on our "solution"
3. Never deliver the project, it's not like they would notice any difference from all other gov IT.
4. Freedom and Profit!!! FTW!
>> "It doesn't matter if they *do* intercept the 'key exchange'. That's the whole point of PKI."
> Are you sure?
Yep.
>> "I send you my public key,"
> But Malcolm, the Man-in-the-Middle, intercepts it. He generates his own public key, and sends
> that on as if it's yours.
That'll have a different fingerprint from the real key. So for this to succeed,
- The client end needs not to be verifying keys against a trusted root OR it is verifying against a trusted root that has been compromised
- The client must ignore fingerprint changes
The above attack has a low probability of success already...
>> "you encrypt a symmetric key with it and send it to me."
> Which Malcolm intercepts, and decrypts using his private key. He now knows the symmetric
> key. He encrypts that symmetric key using your public key, and sends it on to you.
The client has now forged a SSL link with the attacker - *not* the real server. This is important.
>> "The only person that *can* decrypt it is me (with my private key)."
> Which is fine for Malcolm, since he already knows what the symmetric key is.
Yes - but the *real* server will not be able to decrypt the symmetric key - because it is encrypted with the attacker's key, not the real server's.
>> "We now both have the same symmetric keys and can send encrypted data to each other."
> And Malcolm can decrypt all of it, without either of you knowing.
No, Malcolm can only impersonate the real server (assuming the conditions above are met). The real server takes no part in this; it is excluded from the conversation.
So if any part of that conversation requires data held on the real server - say, the contents of your inbox - the impersonation is trivially discovered.
The attack you describe is relatively difficult to effect in most circumstances (requiring the attacker to forge trusted SSL certificates for domains he does not own), and is completely useless against anything but single-ended traffic.
>> "The above is the general routine used in all PKI based systems, SSL/HTTPS included,
>> though there are additional complexities with cert authorities."
> Ah, yes. Handy for guarding against such things as Man-in-the-Middle attacks. But how do you
> know the certificates aren't also being intercepted and tampered with?...
Because even if they are - the behaviour of the system is changed beyond recognition.
So, in summary, for a government to use this sort of vector, it would require undermining the whole security of the web *worldwide*, and would still only catch someone who is truly stupid.
Not that that will stop them trying, of course...
If they are not just stealing public money, they are squandering it.
A couple of viewpoints on this, 1 and this is the important one, no one wants this, it is a negative thing to snoop on communications, communications are better left private, unless the people involved wish to make it public. This is common sense.
2. with projects involving technology make sure the people responsible for allocating the budget actually know the project from the ground up, if they are ignorant of how to do it, then at least they should be taught to step aside and find someone with more knowledge.
This is very much primary school education, we appear to have mentally deficient retards running this country into the ground.
It is abhorrent, we have this marvellous things called technology, it can be used in a utilitarian fashion to produce wonders, to enrich and improve the lives of everyone it touches, what do the government do with it? Try to turn it, unsuccessfully, into some peeping Tom, it is bereft of any virtue.
Won't these oversized gorilla-pig cross breeds just leave, the country has spoken so many times now, only a handful of deluded people are left ticking the wrong box, it is meant to be a democracy, not a hang onto power at all costs tyranny.
It is usually considered best practice to transfer keys physically whenever possible, thereby bypassing the network and any potential man in the middle attackers.
In one example I bring my USB stick over to you and copy my public key onto your PC or device.
Alternatively, a mutually known and trusted intermediary can also act as a key signer who can use their copy of your key to confirm/sign your own copy which you may have have received insecurely via the network.
Another way is to print the key onto paper and mail it via snail mail. Sure, it will be a bit of a pain to hand type but not impossible. OCR can work too, as long as you hand verify it afterwards.
Once you have a signed (known trustworthy) key you are then pretty much set in stone safe.
I've always loved this idea. Nothing would make us more safe from the police and their fellow snoopers than a copy of the entire internet on their hard drives. Personally I'd setup two hotmail accounts just to forward all my spam back and forth just to see whose head explodes from analysing it all.
By the time they'd find anything that would actually criminalise someone that someone would probably have been dead several hundred years.
I do pity the poor helpdesk jockey that will be employed to sift through several million terabytes of data armed with a desktop search engine and pencil. Not sure how long it would take to drive such a person to suicide, but the Facebook accounts of a dozen teenage girls ought to be sufficient.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
