3 ms·
> This might be worth pursuing on a formal basis, particularly through such indicators as financial accounts, email or communications addresses, URLs, etc., ext
by schoen 10y ago
> This might be worth pursuing on a formal basis, particularly through such indicators as financial accounts, email or communications addresses, URLs, etc., extant as canaries which would reveal a data breach.
This is a great idea, but it's not clear how many adversaries would do something in an automated fashion that would give away their existence like this.
Apart from the kind of discredited story about Coventry (https://en.wikipedia.org/wiki/Coventry_Blitz#Coventry_and_Ultra https://en.wikipedia.org/wiki/Coventry_Blitz#Coventry_and_Ul...), I was just reading somewhere again that there was a policy about Ultra that nothing could be done in response to an Ultra intercept without first creating an alternative cover story -- parallel construction, if you like -- about how the Allies knew about it. (See https://en.wikipedia.org/wiki/Ultra#Safeguarding_of_sources https://en.wikipedia.org/wiki/Ultra#Safeguarding_of_sources on this point, although I can't remember where I was just reading about it.)
Automatically contacting or attempting to use accounts or URLs mentioned in random data is probably exactly the kind of thing that wouldn't be done by anybody exercising a similar kind of caution today.
- dredmorbius 10y agoI'm actually fairly certain the fictitious entry concept exists in some shape or form. Even name+company@example.com addresses used for email registration are tells of misuse of registration information. Various honeypot traps utilise similar elements. I'm all but certain financial institutions have at least tracking, if not fictitious entry, accounts seeded within various data lists whose use triggers alerts, though I'm not sure those specifically indicate leaks. Document and video/audio tracking have also been used to determine leak sources. So it's not an original idea. While parallel construction may be used to create an appearance of plausible discovery, an attack against a resource which does not exist would be a strong suggestion that supposedly secret data aren't. I suspect you're overestimating your adversary. Some will act with extreme reserve. But odds are strong that a curious adversary stumbling across what appears to be public URL or IP + service address, and tickling it will happen with some frequency. Other tells could be constructed of a nontechnical nature -- false documents and such. Behavior consistent with access to that information would be an indication. Audits of non-existent financial accounts, inquiries of non-existent companies or people, etc. The hit rate needn't be perfect, but a single tell would be sufficient.
- schoen 10y agoHuh, now I'm thinking Cliff Stoll did something like this in The Cuckoo's Egg with making up a fake project that intruders then wrote to ask for more information about... or am I misremembering?
- dredmorbius 10y agoYes, via Wikipedia: In order to entice the hacker to reveal himself, Stoll set up an elaborate hoax — known today as a honeypot — inventing a fictitious department at LBL that had supposedly been newly formed by an "SDI" contract, also fictitious. When he realized the hacker was particularly interested in the faux SDI entity, he filled the "SDInet" account (operated by the imaginary secretary Barbara Sherwin) with large files full of impressive-sounding bureaucratese. The ploy worked, and the Deutsche Bundespost finally located the hacker at his home in Hanover. The hacker's name was Markus Hess, and he had been engaged for some years in selling the results of his hacking to the Soviet KGB. There was ancillary proof of this when a Hungarian spy contacted the fictitious SDInet at LBL by mail, based on information he could only have obtained through Hess. Apparently this was the KGB's method of double-checking to see if Hess was just making up the information he was selling them. https://en.m.wikipedia.org/wiki/The_Cuckoo%27s_Egg https://en.m.wikipedia.org/wiki/The_Cuckoo%27s_Egg
- dredmorbius 10y agoSee also generally honeytokens: https://en.wikipedia.org/wiki/Honeytoken https://en.wikipedia.org/wiki/Honeytoken http://www.symantec.com/connect/articles/honeytokens-other-honeypot http://www.symantec.com/connect/articles/honeytokens-other-h...