Schneier on Security

Clive Robinson • August 10, 2012 12:54 PM

All jokes aside the incident does raise one point that we see with other systems.

That is he got through an initial check and subsequent checks because of assumptions.

The first assumption was he was part of a family he “tail ended” and after that each subsiquent stage assumed he had been checked by a previous stage so did not check further. Right onto the aircraft where the airline did no checking because they assumed all the previous checks would have stopped anybody who should not have been on the aircraft.

This is a known type of problem however the logical solution which is make less stages and each stage more responsible does not work either because beyond a certain level of checking the checkers will become overloaded and miss checks out.

It is issues like this, that some people are good at spotting and exploiting. But furrther it tells us something important, that security implemented solely by humans against other humans can only be to a certain level, after which it naturaly starts to fail under it’s own weight due to human failings.

This is a known problem on production lines that is usually solved by attaching a “check list” to each item and as they progress down the line each checker stamps/ticks the appropriate box on the check list.

However when the items are sentient an attached list fails because the sentient item can tick their own check boxes.

This is usually solved with livestock handlers and hospitals by attaching a “bar code” or equivelent to the sentient item and the “system” scans the code at each check and the system then holds the “check list” and updates it at each stage.

This taging and scanning is by nomeans foolproof when it comes to sentient items that also whish to deceive which is why we have “tamper proof” seals etc which as we know are either cheap and fallible or expensive and difficult to remove at the end of the. checking process.

After some thought you will realise that checklists with sentient and potentialy deceitful items actuall only works by the consent of the items being checked. And it is this “consent” that actually defines just what level of checking will work in a market driven system.

The follow on from this is an interesting problem that history has not provided a solution to and we end up with the notion of “presumed consent” with punishment for “opting out” with the ultimate threat of forcefull descent when the presumption becomes to onerous. Which is the basis of “democracy” in that we allow ourselves to be governed with the threat of ostracisation from society as punishment, but with the actions of the governing limited by the threat of revolution and “regiside”.

The founding fathers were well aware of this issue and thus to limit the build up of destructive forces and the ensuing great harms to society in general it put significant limitations on the build up of standing armed forces by the governing and the right for the citizens to “bear arms” and in return requiring them to be responsible for their own defense by the raising of malitia.

Unfortunatly as a solution it has failed the US now has the worlds greatest standing armed forces (estimated to be twenty times that of the next nearest nation) and “Law enforcment agencies” without liability (DHS/TSA) to the people. All that has changed is “The King” is now “A President” and Barons and Earls have been replaced with corperations and politicians.

As was once engraved by diamond ring on glass pain “The more things change the more they stay the same”.

Clive Robinson • August 12, 2012 2:22 AM

@ Anon5,

I’m not sure where you’re getting your information, but China has a larger military than the US

Dirk was quoting a part of what I had said a little further up this page. The problem with “estimates” is what you count and why and do you do it equally for both sides (usually not). This issue was something that came up during the “cold war” time and time again one way or another.

For instance in Israel and Switzerland and a number of other places what would normly be considered to be “civilians” could be counted as “soldiers” because of the “minute man principle”

Likewise the US coastguard, would by many other nations be considered part of the US standing army along with the “weekend warriors” and the various armed LEO’s.

It is argued that the “War Hawk” estimates for Chinese armed forces are usually “significantly inflated” by these “internal security” organisations of which China has many. Likewise the armed forces and LEO’s of the “buffer” or “vassal” nations such as North Korea around China’s borders (during the “cold war” the US military pretended they stood alone for their estimate whilst including all Warsaw Pact nations into the Russia estimate).

And this is just a “head count” it does not go on to include the “force multipliers” of the weapons and other technology available to be deployed by these bodies able or not (with some technology like drones you could be significantly disabled such as have no legs and still be effectivly a soldier, sailor or airman even though no where near the “front line”).

Another trick used during the cold war was “reliability of weapons”. It was known that US nukes were (initialy) due to all the protection mechanisms of PALs not as reliable as those without by a factor of about two. Also the US ICBMs were also being rotated through maintainance rather quickly as well and some estimates indicated as much as a third were unavailable at any one time. However the argument was that this was not the same for the Warsaw Pact (based on their short range tactical weapons). Thus after much hand waving the US military strove to have few tactical nukes and tried to maintain atleast two strategic nukes against the highest estimate of all Warsaw Pact tactical and strategic nukes combined, not just the “Russian ICBM’s”

So you have to take “the estimates” with a pich of salt around the size of “Lot’s Wife”.