Would not a better way to do this be to take the spammer's approach? Spam has gotten to be quite good at sounding realistic. It's also normally quite naive. But instead of emailing real people, just setup a vast network of bot addresses that all mail each other in realistic ways. The content of the messages can be algorithmically generated to appear like nefarious conversations.
I'd love to see NSA and FBI members burning money and time tracking down spam bots.
> Would not a better way to do this be to take the spammer's approach
On the other hand, anti-spam has gotten very good at distinguishing wanted and unwanted content.
If Google can do a reasonable job binning email spam, we must assume the NSA can do a reasonable job ignoring even relatively sophisticated keyword spamming.
> If Google can do a reasonable job binning email spam, we must assume the NSA can do a reasonable job ignoring even relatively sophisticated keyword spamming.
Indeed; however the recent Snowden leaks have given us a replacement - we know that the NSA stores indefinitely any encrypted messages. So, simply include inline a random message. Boom, you've permanently used up that much of their space for nothing.
I was thinking a new English lect, or a form of ebonics is needed that incorporates "bad" words that set off filters. Once that becomes the vernacular it will be very difficult to figure out what's useful and what's not. To be forced to do this however is stupid. Legitimate intelligence is still reasonable. But we're so far beyond that that it has come to this.
I'd love to see NSA and FBI members burning money and time tracking down spam bots.