Strategic Security Intelligence

Information Warfare

DRAG Project - Fred's Review

Copyright(c), 2005 - Fred Cohen - All Rights Reserved

The Setup

Ever since the late 1990s when folks started to use control over their forums in conjunction with select planted identities and people to generate enhanced market around their Internet-based products and services, I have been interested in this phenomena. I have avoided its use for personal gain because I consider it to be a deception and therefore unethical as an approach to sales and marketing. The world has, of course, continued to do it, and at some point I knew that an experiment was warranted. It is not likely the first of its kind nor is it likely to be the last.

Students in my graduate class in deception, perception management, and propaganda (it had a longer title naming all sorts of other related items but the Dean wisely shortened the title so it would fit into the catalog) were nearing the end of the UNH trimester and their attempts to generate projects for their required practical use of the material in the class were failing. So I offered them an alternative. They bit.

The overall effort was a short and therefore forceful perception management exercise. Despite the short time frame and the lack of subtlety it was necessary to infuse attack group members into the iwar forum rather quickly and only short-term measurements were able to be made. Ethical considerations require notice of group members of what was done, so the long-term effects could not be measured.

The material used as the core subject matter ("Did China Win the War?") was a paper that I wrote and posted to the iwar forum but that got little or no prior attention from the forum. This made it an ideal for the perception management effort for the following reasons:

So I got permission from the Dean at the school to make sure it was fine all around, set up a sufficiently controlled condition for the experiment, and even asked the forum members if they would like to play a game - which generated no negative responses and some mild positive response. Off we went.

The Action

The action started with the infusion of team members into the iwar forum. As a forum in which all group (but not all personal) communications could be observed and were automatically recorded, this was considered to be a nearly ideal proximate situation. Proximate techniques were therefore selected over other techniques. It was considered important to the team that they be considered as members of the group, but generating acceptance was problematic because of time scales, and because of the relatively low volume of the group, coming in with lots of messages from a large group of new people was considered too risky up front. Identities were established in the first week of effort, users were added to the group through the normal method, with me as the moderator asking the same questions I ask of all potential group members - enough to make sure they are real people and not just spammers. Because of the lack of time the identities were relatively thin, but adequate in depth for the level required for this forum.

At some point, one of the team members introduced the article about China, which was perhaps a bit too bold a maneuver, but which did not have a serious negative effect and got the information out there. Uptake was slow and little targeting of specific individuals was done at this point. We decided to pick up the focus a bit.

Intelligence efforts were undertaken to identify and target specific individuals in the group. This included looking through the iwar forum history to identify the thought leaders and group movers, looking up general and specific information on them, and identifying behavioral patterns associated with those individuals. This was specifically targeted toward finding behavioral mechanisms that could be exploited to cause them to post to the forum with an opinion in the desired direction - or just to keep certain topics alive. The result was ultimately a weak mapping between individuals and mechanisms that could be exploited against them.

Using these mechanisms, the team started in at a higher pace to generate specific responses to have the desired effect. Targets were exploited slowly at first and with increasing rates later on, using the basic principle of moving the group through cognitive resonance where possible and dissonance when specific responses called for it. The rate increase was very substantial over time, but it was controlled so as to not be so large that it caused the group to collapse or sway from its norming and performing mode. The desire was to keep the group from storming and as moderator this is generally my desire - although we want vigorous conversations and welcome all opinions, when to gets to personal comments, I push back gently. This only happened once.

Several very specific goals were achieved. The most obvious ones were getting several forum members to essentially buy into the whole premise of the paper, getting increased levels of discussion around the topic, generating interest from news media in the material (without contacting them directly except for forum members who happen to be in that community), preventing the direct exposure of the deception or the team members, and using the material from the course to demonstrate in a practical way to the students that the specific mechanisms really work as advertised (or perhaps to change the advertisement from what we learn).

As the effort neared its end, a decision was made to let it die out on its own by no longer supporting the discussion with ongoing impulses of content from team members. This was out of a desire to return to steady state for measurement purposes as well as to give the team time to write up their results before grades were due.

Finally, just after the final paper was handed in and grades given, I posted the news to the iwar forum so that members could enjoy the fruits of their labors and hard feelings, if any, could be hashed out directly in an open and pleasant manner to mitigate any possible harm to the targets of the deceptions.

The Report

The report is now posted and permanently available on the iwar Web site. Over time I will try to get the full details posted, but we will see. My view of the report is that, for a Masters level class, and given the time available to do the whole process and write up the report, it shows a lot of knowledge gleaned from the course and practically applied. It also contains the requisite amount of bowing to the professor's wishes and attempts at making the professor feel good about himself for most professors to give it a decent grade, even though I count it as a negative and thinly veiled, especially in a course such as this.

The report was lacking in several important ways (although a lot of it is covered in the appendices). One really important thing it missed was the overall context. Another is that it did not get to the real issues and use adequate metrics for good measurements. Time frames for graphs should have been weeks instead of months in order to show a project of this duration and makeup. It also greatly exaggerates the brilliance of the students by sharpening their successes and leveling their failures as well as ignoring the fine work that forms the basis of their success - to wit all of the citations associated with all of the techniques they applied.

If I only had Ph.D. students and funding to do the job properly, we could probably generate really good metrics for doing perception management in Internet forums (and make me a small fortune along the way), but what am I talking about - the marketing firms must be in this game big time by now...

The Aftermath

It was my great good fortune to have done this experiment in the iwar forum. The forum members were generally most gracious about the effort, some enjoying a bit of a laugh, and hopefully most thinking about it fondly, if at all, over their Winter Solstice break (let's face it - that's what it is - the Pagans have won in the long run). The next time I ask them if they want to play a game, they may perhaps answer "No" more resoundingly, but you never know.