[Dailydave] information operations efforts and data carving

[Jukka Ruohonen]

On Thu, Aug 09, 2018 at 01:09:57PM -0400, Dave Aitel wrote:
>    Clemson:Â [2]https://www.cyxtera.com/blog/data-carving-the-internet-res
>    earch-agency-tweets

This was a good take on things. I generally also applaud the constructive
criticism instead of the ranting strategy...

>    various curves if you're ignoring the content. Using NLP techniques to
>    analyze the tweets at various points helps us make better conjectures
>    about the flow of information through an adversaries IO systems.

But it is still social media. Now I've seen quite a few papers recently
about vulnerabilities viz. Twitter.  Some of these are relevant; there have
been some information leakages about things I consider relevant myself
(i.e., open source). But now people are attaching the "zero-day" label to
their papers, which kind of makes you wonder even about correlations.

>    But if you read our previous work (linked above) we've also seen bot
>    accounts that are virtually indistinguishable from real people until
>    you go try to verify them on LinkedIn. We've also found massive
>    client-side exploitation networks. 

But have you considered the possibility that it is you that is being fooled?
What I mean is that a big attack vector in the future will likely be machine
learning and artificial intelligence. People will find ways to poison your
training data with false positives, outliers, and whatnot. In fact, I'd bet
that many are already doing this.

Big issues. Big questions. I think EFF did a reasonable job a couple of days ago:

https://www.eff.org/wp/cautious-path-strategic-advantage-how-militaries-should-plan-ai

- Jukka