When you read the below you’ll probably just flash the thought: “What a bunch of fucking mother fuckers!” 1
And you would be correct.
FOIA requests to the FBI are processed by searching the Automated Case Support system (ACS), a software program that celebrates its 21st birthday this year.
Not only are the records indexed by ACS inadequate, Shapiro told the Guardian, but the FBI refuses to search the full text of those records as a matter of policy. When few or no records are returned, Shapiro said, the FBI effectively responds “sorry, we tried” without making use of the much more sophisticated search tools at the disposal of internal requestors.
“The FBI’s assertion is akin to suggesting that a search of a limited and arbitrarily produced card catalogue at a vast library is as likely to locate book pages containing a specified search term as a full text search of database containing digitized versions of all the books in that library,” Shapiro said.
The DoJ has contended to Shapiro and others that only one of ACS’s three search functions, the Universal Name Index (UNI), is necessary to fulfill the law. The UNI search does not include the text of the files in the ACS, merely search terms entered — or not — by the FBI agent handling the case in question.
Shapiro told the Guardian that the reason the DoJ gave for refusing to use its $425m Sentinel software to process FOIA requests after ACS had failed to recover records was that a Sentinel search “would be needlessly duplicative of the FBI’s default ACS UNI index-based searches and wasteful of Bureau resources”. […]
The FBI’s chief technology officer during the second George W Bush administration, Jack Israel, […] said four years ago. “It’s based on an IBM mainframe with legacy database and programming technology, and I would say one of the main things that strikes you as a user of ACS is that you’re dealing with the old IBM green screens. You’re not dealing with a web-based environment, which everyone is used to from the internet.”