On ScareMail Generator (Analysis)

Justin George

ScareMail Generator: Analysis

In the spirit of Halloween, I’ve decided to cover the ScareMail Generator: A word generator that processes various words and phrases into an incoherent horror story that tends to be slightly more stupid than scary, likely to be shared by hundreds of clueless aunts and uncles as something genuinely frightening for the sheer horror that such words could be chained together in such a way. But the intent of the ScareMail Generator, is to address a horror that is not just limited to Halloween, but even all year-round.

In recent years, the NSA has taken to identifying certain “keywords” that are supposedly used to detect terrorist communication and behavior. Words like “plot”, “facility”, even “packages” have become no longer considered acceptable in normal written text, but taboo as the name “Voldemort” in the Harry Potter series.

While in theory this is a precautionary measure that can be said to be done for the greater good of protecting the free world in a post 9/11 society, the author describes it as closer to “a governmental surveillance machine run amok, algorithmically collecting and searching our digital communications”. In a present time where our constitutional rights (and more than often the rights themselves) are challenged daily, ScareMail Generator presents itself as a bot with a very important message behind its nonsensical English text: “words do not equal intent”.



The source text for all of the stories is Fahrenheit 451, a personal favorite and a cautionary tale about the degradation of society via modern conveniences and vanity, but also about censorship. This is likely to be intentional, as most other blocks of text could perhaps accomplish the same task; the selected text in this case is done either for a symbolic effect or for the purpose of having a format that resembles a narrative more than anything. The protagonist “Montag” or side character “Clarisse” show up from time to time in with my generated text, a reminder of the source material.

The author, Benjamin Grosser, has a history presenting at events related to countering projects like the NSA’s programs, indeed ScareMail itself had been first revealed at PRISM Breakup in 2013.  The fact that the source code and inner workings of ScareMail are freely available to the public further enforce the idea that just like the thing it is working to counter, ScareMail has nothing to hide.

The purpose behind ScareMail is part obstruction, part demonstration, and wholly to ensure that NSA programs like PRISM and XKeyscore don’t really have a clue when it comes to looking up “trigger” words. These programs work off of finding these blacklisted words through loads of read (typically without your permission) emails and building a record off of it. But if the programs are overloaded with junk examples of those words being used, like the types of narratives that ScareMail produces, the programs and their databases become inherently  worthless. Freedom of speech has been often contested in the name of preventing terrorism, but ScareMail doesn’t try to convince you that NSA surveillance is in the wrong here; just that their programs are poor implementations in the name of that security.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s