If a used has a password on a system that is 28 ASCII characters, on a system, lets say it's my.gov.au and then a few years ago a flaw is discovered which limits passwords to 20 characters and the user now discovers that the password for the site is the original 28 character password truncated at 20 characters, is it reasonable to conclude that the password is stored in the plain, and how could one make this argument to the someone with a legal rather than an information theory background (ie. hashing is one way etc).
How can I convince a legal person that when a password it is possible that it is stored in the plain?
3$\begingroup$ I don't necessarily agree there. It's entirely possible that they truncate the password before hashing it, for update or comparison purposes. The fact that they do truncate it in isolation does not prove that they store passwords in plaintext. $\endgroup$– MorrolanNov 3, 2021 at 12:05
$\begingroup$ Do you really want to convince them that "it's possible"? That seems like a very low bar. Someone who is not technical might not even realize there is another alternative. $\endgroup$– bmm6oNov 3, 2021 at 16:52
I'm a technical person, and understand the argument being made in the question. But before I start to side with these arguments to assert the insecurity of the password storage system, past or present, I'd need to be convinced that when 28 ASCII characters was accepted, a password typed with an error past the 20th character was refused. If not, it's entirely possible that since the origin only the first 20 characters of the password have been significant, and that is limit now enforced, and for the rest the password was and is stored properly password-hashed.
Even then, I won't be convinced "that the password is stored in the plain". It's entirely possible that he password now is stored properly password-hashed, with the conversion from an old format to a new one when the password, or it's first 20 characters, are first used in the new system. Such seamless conversion from legacy to new password format is standard procedure.
I won't even be convinced that the password was stored in the plain. It's entirely possible that the password was stored encrypted with a reversible encryption and some secret key. That would not be properly password-hashed, but still better than "in the plain" (and even quite satisfactory if the password decryption and handling occurs only in a secure environment, like an HSM).
I'm not saying it's impossible to use "Jedi mind tricks" and convince the legal person of something, as was done for the certification of a plane. But that's something I do not condone. Beside, it's off-topic.