e-Discovery: How mandatory are the data fields requested by the Congress?

Question

Some data requests sent by the US Congress sound like they require some metadata fields to be populated attached to files, “unmodified”.

For example, this request (as do other similar ones) says:

Responding to Oversight Committee Document Requests

5. Documents produced in electronic format should be organized, identified, and indexed electronically.

6. Electronic document productions should be prepared according to the following standards:

   d. All electronic documents produced to the Committee should include the following fields of metadata specific to each document, and no modifications should be made to the original metadata:

   BEGDOC, ENDDOC, TEXT, BEGATTACH, ENDATTACH, PAGECOUNT, CUSTODIAN, RECORDTYPE, DATE, TIME, SENTDATE, SENTTIME, BEGINDATE, BEGINTIME, ENDDATE, ENDTIME, AUTHOR, FROM, CC, TO, BCC, SUBJECT, TITLE, FILENAME, FILEEXT, FILESIZE, DATECREATED, TIMECREATED, DATELASTMOD, TIMELASTMOD, INTMSGID, INTMSGHEADER, NATIVELINK, INTFILPATH, EXCEPTION, BEGATTACH.

Assuming this is sent to a small company, it is quite possible the documents are stored on employees’ work (or even personal) computers, possibly as just Word documents. It is unlikely that they are organized in a database that would contain these fields.

These fields seem to be mentioned by a discovery service provider company, but I have not been quickly able to understand where these fields come from, or why Congress seems to have an expectation that the respondent will have the documents organized with these fields.

I’m neither a legal professional or from the US, so out of personal curiosity (and professional curiosity in the field of IT), I am trying to understand:

• Is there is an actual standard that proscribes these fields?

• Is there an actual federal US legal requirement to either maintain documents with this metadata (or at least generate it on export)?

• If this applies only to businesses of certain size, what is the threshold?

That is: do I need to worry about storing small business or large corporation’s documents centrally and annotated with these fields?

Answer 1

The reasonable implication and interpretation of this is that metadata on a file to which a field is not applicable is not required.

For example, "FROM, CC, TO, BCC" while applicable to email, would not be applicable to a spreadsheet that was never shared with anyone for purposes of making a calculation. Unmodified metadata can only be provided to the extent it exists.

Answer 2

These come from an ESI discovery protocol, which as they say

is a mutually agreed upon plan on how both sides will access data and, eventually, make it available to the other side

though there is probably no mutuality in the case of Congressional discovery. It's unclear whether there is a single authoritative standard, but here is an example (in appendix 2.1). The field "INTMSGID" is not defined there, so you would probably have to hire a lawyer to find out what they want. The list you cited frequently pops up in letters from Congress to federal agencies, and may be mandated in some federal record-keeping regulation. If Congress subpoenas you and your documents, they may be less presumptive and demanding.