Slashdot Mirror
US Gov't Makes a Mess of Classifying Sensitive Data
coondoggie writes "Protecting and classifying sensitive information such as social security numbers shouldn't be that hard, but (perhaps not surprisingly) the US government has elevated complicating that task to an art form. It seems that designating, safeguarding, and disseminating such important information involves over 100 unique markings and at least 130 different labeling or handling routines, reflecting a disjointed, inconsistent, and unpredictable system for protecting, sharing, and disclosing sensitive information."
This was the conclusion of a recent report (PDF) by the Government Accountability Office, which also "found areas where sensitive information is not fully safeguarded and thus may
remain at risk of unauthorized disclosure or misuse."
If you closely tag it to everything you do, you're doing it wrong. Unless they are a financial institution, tell em to shove it. Hell, it took my university until 2004 to figure out not to use that as a student ID number and encoded (without encryption) in the magstrip of the ID cards. Most places will allow you to get credit from them (like utilities) without it... if you ask.
Having read the article, and being a US Gov't employee, let me just say that Cooney has unnecessarily confused the issue. Some of the 50 examples he lists are duplicates ("1. SENSITIVE", "17. SENSITIVE (SENS)", "40. SENSITIVE BUT UNCLASSIFIED (SBU)" are all the same thing, as are "3. SBU-NF" and "4. SBU/ NOFORN", and several others). Many of the others are mixing apples and oranges. Items 5-9 deal with the data ownership, which is reasonably treated differently from "15. SOURCE SELECTION SENSITIVE" or "33. ATTORNEY CLIENT" information. Is the list Cooney presents absurd? Possibly. Could the Gov't marking system be simplified? Probably. But don't do it on the basis of this article.