System Security for Cyborgs

ecksor

Short paper by Ross Anderson. PDF format, only 3 pages, interesting overview.

http://www.cl.cam.ac.uk/~rja14/Papers/cyborg.pdf

Martyr

This reminds me of what that guy Kevin Warwick does..
http://www.kevinwarwick.org/

I can see the benefits of using EEPROMS in humans for such things
as storing medical details and for sure, developing anatomic devices
for those without legs and arms for example.

Even artificial eyes and hearing aids when the technology becomes available is good news for some.

But anything that communicates with something on the outside or that can be interfered with externally, wouldn't make me very comfortable.

Like that 'subtle' heart-rate monitor to stop your phone ringing..if you're
cycling..haha, thats bit silly, but who am i to say.

Gilgamesh

really seems relevant on putting this cr@p in the security forum, after what I read in the 'read this first' thing.

Oooohhhh!!! look at me, ... the Cyborgs are coming,...... which port do I block to refuse him access to my house

ecksor

Did you actually read that paper? Please explain how it isn't relevant.

If you're not interested, then don't post. If you have a problem with the forum rules, then post on the relevant thread.

Gilgamesh

The entire document is a total 'What if' scare, seriously, so you expect to have a Wifi Card shoved in your head so someone can hack your brains and make your heart explode.
Even is you just have an interface in your arse, I am sure that someone will notice if someone would want to 'interface' with it.

I understood that that this forum is for actual issue, going from how to setup a firewal or how to block adware properly, not about Apocalyptic theories how someone can be killed in the future by overclocking his heart.

In 20 years time, if someone is haviong problems that Adware is been loaded into his eye implant, then that, imho would be an issue for this forum.

Chalk

i think you should read the stickies
thats the exact opposite of what this forum is for.

Gilgamesh

Chalk wrote:

i think you should read the stickies
thats the exact opposite of what this forum is for.

okay, so it isn't a tech support forum, and the adware for the eyeballs wouldn't be allowed to be put up here.
but do you really think considering what is going to happen when we get Cyber implants in this day and age is relevant to a computer security forum, I honestly don't and would have thought this might even be better in After hours or something like that

ecksor

Gilgamesh wrote:

The entire document is a total 'What if' scare, seriously, so you expect to have a Wifi Card shoved in your head so someone can hack your brains and make your heart explode.

Ah, so you do claim to have read it? Odd since you seemed to think it was about defending against cyborgs instead of defending risks to cybernetic implants. (I can't see the mention of Wifi cards there).

Quite a lot of computer security research has been conjectural in nature (would it be research otherwise?), but that hasn't made it irrelevant. Buffer overflows were considered within the realms of research before they became a common problem for example.

I understood that that this forum is for actual issue, going from how to setup a firewal or how to block adware properly, not about Apocalyptic theories how someone can be killed in the future by overclocking his heart.

So what makes you think that pacemakers (for example) won't be susceptible to outside interference or information leakage? Information has been shown to be leaked from all sorts of unlikely sources in the past. Medical data is a frequently targeted information source also, and medical technology is one of the main (perhaps the main going by some of the cybernetics students I know) applications of cybernetics technology.

You may think that this is very far off in the future, but this technology is being developed now and the implications need to be considered now, not later when it is too late to change design flaws (as has happened with so many other things in the past).

Gilgamesh

okay,
I guess I am interpreting the rules of this forum incorrectly.
Fairnuff, will humbly bow in front of you Ecksor.

Now to the note.

In this case even my first statement was relevant, as you put it about fending off cyborgs, who said it won't be possible to hack peoples brains then and make them do what you wan't?
Allthough this is much further in the future, it would also be possible.
So it would be a concern relevant to this link.

Cybernetics is coming, and I totally agree with that in all ways, but on the other side.

I do beleive that the damage that could be caused to human implants can be easily limited, by not making the monitoring equipment or the implants, remote readable, but only with direct interface into the unit.
through that, it will be probably even uninteresting for people to go for the Cybernetics, than actually just kill someone, takes less resources and is much easier than plugging something into someone to damage them or kill them

I think the bigger issue of Cybernetics is going to be the reliability of the Software it is running on, as a slight 'Bug' would or could be fatal.

Anything the human makes, will have weaknesses and there will allways be people exploiting these weaknesses.

ecksor

Gilgamesh wrote:

In this case even my first statement was relevant, as you put it about fending off cyborgs, who said it won't be possible to hack peoples brains then and make them do what you wan't?

Nobody at all, but the paper doesn't address that IIRC, which was all I was saying.

I do beleive that the damage that could be caused to human implants can be easily limited, by not making the monitoring equipment or the implants, remote readable, but only with direct interface into the unit.

That's a bit vague. Are there no emissions from such devices?

through that, it will be probably even uninteresting for people to go for the Cybernetics, than actually just kill someone, takes less resources and is much easier than plugging something into someone to damage them or kill them

I didn't see murder listed as a threat under consideration at all in the paper actually (apart from a reference to sci-fi writers). If murder is what you are considering, then perhaps you are right.

I think the bigger issue of Cybernetics is going to be the reliability of the Software it is running on, as a slight 'Bug' would or could be fatal.

Anything the human makes, will have weaknesses and there will allways be people exploiting these weaknesses.

It's hardly a bigger issue, it's included in dependability, it isn't separate to it.

3DES

Gilgamesh wrote:

I understood that that this forum is for actual issue, going from how to setup a firewal or how to block adware properly, not about Apocalyptic theories how someone can be killed in the future by overclocking his heart.

I would consider security in cybernetic devices to be on-topic, as would discussing the security aspects of wireless sensor networks, neither of which have anything to do with application layer problems such as setting up a firewall or blocking spyware.

On to the paper: I was disappointed to find that practically nothing on the actually security aspects of medical sensor implants was actually covered. At times I found myself wondering if I was actually reading a paper about the motor industry, as the author seems far more knowledgeable in that than he is in security issues. The paper seems to talk about everything except the actual security issues related to medical implants, going off in tangents unrelated to medical implant security, at one point even mentioning printers that refuse third-party ink cartridges.

For instance he says:
"What about the patient who’s forgotten to pay the software license renewal on her pacemaker?".

This has nothing to do with security and would likely not even be an issue in the first place.

And also:
"If I were advising a gifted security research student who was looking for a thesis topic in this general area, I might well suggest usability and maintainability as the most important – and the most challenging – of the problems."

It appears that author has completely missed the point of security research if he believes that "usability" and "maintainability" would be "the most important" points to be covered in a security related research thesis.

There was no discussion on the real issues that might be encountered, such as network topology, node capture, authentication or protocols etc. He even failed to mention what possibile transmission media these devices might use to communicate.

Overall it read like a paper written by someone who has never studied or researched any aspect of network security.

ecksor

3DES wrote:

On to the paper: I was disappointed to find that practically nothing on the actually security aspects of medical sensor implants was actually covered. At times I found myself wondering if I was actually reading a paper about the motor industry, as the author seems far more knowledgeable in that than he is in security issues. The paper seems to talk about everything except the actual security issues related to medical implants, going off in tangents unrelated to medical implant security, at one point even mentioning printers that refuse third-party ink cartridges.

I'd be careful before dismissing Ross Anderson as someone who didn't know what he was talking about.

Motor industry analogies have been around for years for assessing the state of computer security, sometimes because it appears that current trends wrt standardisation and risk mitigation are similar to that industry's state approximately 70 to 80 years ago and sometimes because it's a good example of a ubiquitous technology where various risks are accepted or dealt with and where attitudes to those risks have changed dramatically over the course of its existence.

For instance he says:
"What about the patient who’s forgotten to pay the software license renewal on her pacemaker?".

This has nothing to do with security and would likely not even be an issue in the first place.

It would be if someone was basically being held to ransom by the person licensing the software, as he suggests in the previous sentences, no? If someone could stop my pacemaker because I owed them money then I
wouldn't feel very secure. If they can stop it for that reason, what other situation would lead to my safety being compromised? Why would it likely not be an issue in the first place? If such scenarios aren't considered now with appropriate public policy put in place to prevent them then I don't see why it might not happen. Consider the US and their mix of a mercenary health care system and lobbies that get things like DMCA enforced in some crazy ways and it doesn't seem far fetched.

And also:
"If I were advising a gifted security research student who was looking for a thesis topic in this general area, I might well suggest usability and maintainability as the most important – and the most challenging – of the problems."

It appears that author has completely missed the point of security research if he believes that "usability" and "maintainability" would be "the most important" points to be covered in a security related research thesis.

He explains in the paper how usability is related to risk, so I think you should cover your statement in relation to that if you're going to make such statements. He also explains why he believes that maintainability is important. Do you not consider that remote software upgrades to a critical piece of hardware that your life depends on is an important security problem?

There was no discussion on the real issues that might be encountered, such as network topology, node capture, authentication or protocols etc. He even failed to mention what possibile transmission media these devices might use to communicate.

Overall it read like a paper written by someone who has never studied or researched any aspect of network security.

It looks to me as if you want some sort of cookbook on how to secure the body against prevalent threats or were you hoping for something else? Are you familiar with much of the author's work?

3DES

ecksor wrote:

It looks to me as if you want some sort of cookbook on how to secure the body against prevalent threats or were you hoping for something else? Are you familiar with much of the author's work?

Perhaps I'm being too hard on the author; coming from an engineering background I'm more concerned with the technical aspects of the security of such devices as opposed to the social policy aspects of security that he focues on in this paper.

ecksor

The paper isn't supposed to solve any practical problems, he's trying to figure out what the problems will be.

Still curious from your comments about how familiar you are with his work though, he has solved umpteen practical problems in his time. If you're working on security from an engineering point of view then you should have read his book on the subject IMO: http://www.cl.cam.ac.uk/~rja14/book.html

3DES

ecksor wrote:

Still curious from your comments about how familiar you are with his work though, he has solved umpteen practical problems in his time. If you're working on security from an engineering point of view then you should have read his book on the subject IMO: http://www.cl.cam.ac.uk/~rja14/book.html

I had not actually heard of him until I read this thread. My interests in security come from data communications and telecommunications, mostly on the data-link and network layers. In other words topics such as medium access control and secure routing, but also encryption algorithms such as AES and public key cryptography. Also from doing an internetwork security module in DCU I've become interested in security in wireless sensor networks (a field dominated by Adrian Perrig if you want to look up papers); techniques used in WSNs in the physical layer could be applicable to medical implants/cybernetic devices of the future, although the network topology would differ considerably.

His book appears to be in the library, so I'll definately check it out.

ecksor

Heh, the first paper I downloaded from Perrig's website cites Anderson twice.

Capt'n Midnight

Know attacks against smart cards - PDF 346KB Many of the remote techniques in this could also be used against embedded systems. Defense forces world wide have invested in non-lethal crowd control, one system involves >100GHz radiation to create a burning sensation in the skin, others use low frequency sound, embedded devices could interact with these in unknown ways.

There is litterly a lifetime to hack in to the system.

With airliners techniques like using old, stable and relatively simple CPU's is used (plural so they can cross check each other) and software testing is given a real budget. They've avoided the intermittant heat caused 386 multiply bug and the far less serious pentium flaw buy using CPU's that you could be resonably sure had as far as possible every possible state tested before being used. I'd doubt that will happen with cyborgs , seeing as how speed at low power would require cutting edge (read not fully tested yet) technology.

Love that Far Side Cartoon where a guy on a plane using a laptop gets a message.. "New Hardware Found , Airbus 310, please wait your device is configured"

Stabshauptmann

What I wouldnt give for some extra memory
The thought of police imobolizing ppl is just downright scary.
Breaking news: 10,000 Protestors imobilised today in dublin...