[Cryptography] Signal hypothetical use case becomes practical since cellular providers have started censoring private text messages

Ray Dillinger bear at sonic.net
Fri Dec 31 21:25:12 EST 2021



On 12/31/21 23:47, Whitfield Diffie wrote:
> > That link points to a site that contains false information
> > that has the potential to quite literally kill thousands
> > upon thousands of people if acted upon.  Personally,I
> > think it’s a good thing that the telecoms are censoring it.
>
>     I am shocked to learn that they are allowed to censor anything. 
> This is a personal message, not a broadcast.  Are they allowed to
> censor email?  Telephone conversation?
>
>                             Whit
>


There is no distinction between 'broadcast' and not, any more.  It's not
like mail servers will otherwise routinely deliver to
"all at somedomain.org"  now the way they did in the pre-spam era. After
the "all" email addresses were all blocked, it was easy to spot millions
of copies of something.  So people started blocking messages if the
server had seen a hundred copies of it already. So spam moved away from
being identical to other spam.  Sentence structures are rearranged,
words are substituted, paragraphs swapped around, sentences transformed
by programs that parse and deparse the original, etc.

So now the spammer is sending millions of "personal messages," no two
alike.  And, usually, sending it from a botnet, where it comes from
thousands of locations whose users are also sending legitimate email.

And yeah, those are routinely blocked. 

Surely you've noticed the tortured grammar, word misuse, misspellings,
and "wrong" phrasings in the spam for the last at-least-two-decades? 
That doesn't all happen because this is an ESL speaker somewhere in a
faraway impoverished country whose education has failed them (or who has
failed their education) but who's trying to write these messages in
English anyway.

Nor, to invoke an oft-repeated claim, does it happen because the spammer
is administering an adversarial "intelligence test" in order to filter
out smart people who'll notice the language failures.

Bear



More information about the cryptography mailing list