Using Artificial Intelligence to Help Clear the Investigation Backlog

I think that a better solution, or a great addition to this solution, is going to be an update to the guidelines used by adjudicators in making their initial decision. In particular, I think that cases that include derogatory financial information need to be looked at differently.

If appears that the current methodology ignores (or pays too little attention to) mitigating factors provided as part of the initial application or were provided during the investigation.

In my case, my wife and I suffered through some serious financial issues caused by a failed business, the housing crisis of 2007/8, and the poor economy of the following years that resulted in several periods of un- and under-employment for me. I was up front about everything when I applied for my clearance in 2016. I provided explanations (as well as you can squeeze them into the SF86) and provided far more detailed information to my investigator in early 2017. When I received my SOR, I replied with more and updated information and was still forced to a hearing. During the hearing process I duplicated all of the information already provided and the judge ruled in my favor in about a month. DODCAF threatened to appeal that decision but decided against it about a week later.

The intelligence needed, artificial or natural, is for adjudicators to look at cases that have been decided by the ALJs and incorporate that into their decisions. If it can be shown that the applicant is very likely to be granted clearance after a hearing, they should be granting clearance at their level.

This would significantly reduce the number of cases going to hearings and would help with the workload on the lawyers and the ALJs. It could have cut at least six months from my clearance process and I don’t think that it would have an affect on the rigorousness of the overall process.

3 Likes