In response to John:
<There's a case called Public Citizen v. Carlin that explicitly rejects the
argument that metadata must be preserved.>
The Sedona Guidelines are all about reasonableness in recordkeeping
practices as a response to unreasonable spoliation challenges and charges
like what must have come about in the 1999 case mentioned above. But the
above ruling, and Sedona, doesn't mean that metadata preservation is not
important. It does contribute to evidentiary value of records, especially
where computer-generated audit trails help trace chain of custody like in
J.'s issue about migrating MS Outlook content between repositories. There
are 3 much-cited cases (below) which went the opposite way from Public
Citizen v. Carlin.
In response to Chris:
<While it would be my recomendation that all metadata be retained and be
migrated with the record, I don't think ther eis a legal requirement to do
so.>
Here in Ontario, the Evidence Act, R.S.O. 1990, c. E.23, s. 34.1(5.1)(7)
specifies audit trails as a means of proving system reliability and
integrity which is necessary to establish authenticity which in turn is
necessary to establish relevance which is required for admissibility.
Case law is useful if we're talking "legal requirements" . Here's what I've
taken away: Case law and legal commentators in Canada and the US have made
it clear that anyone seeking to admit an electronic record must lay a proper
foundation for reliability. It's also a Canadian standard CAN/CGSG 72.34
2005 Electronic Records as Documentary Evidence, written by the federal
government standards board. I consider repeated rulings on the same issue
to be pretty strong indication of a "requirement" and something you would
want to mention in a risk management report or guideline - especially if it
justifies your job and your efforts. In these 2 cases courts refused to
admit evidence because there was no proof of system reliability:
R. v. Sheppard (1992), 97 Nfld & P.E.I.R. 144 (Nfld. S.C.);
R. v. Rowbotham (1977), 33 CCC (2d) 411 at 416 (Ont. Co. Ct.)
In the US case People v. Holowko, (1985), 486 N.E.2d at 878-79 the fact that
records were computer-generated made the evidence admissible (because it was
reliable not because it was a business record exception to the hearsay
rule). Presumably the FBI proved they were computer-generated. Audit trail
reports indicate whether something is human-authored or computer-authored.
Audit trails that are automated can also prove that a record was created in
an ERMS contemporaneous with the activity it is evidence of (as opposed to
being created after the fact when litigation was anticipated -scrambling to
create exculpatory evidence and all that). This contemporaneous rule is
actually legislated: Evidence Act, R.S.O. 1990, c. E.23, s. 33(3) . Also
there's a US case which upheld this rule: United States v. Blackburn, 992
F.2d 666 (7th Cir. 1993)
The Canadian Standard CAN/CGSG 72.34 2005 specifies in section 8 "audit
trail" what audit metadata needs to describe, how it should be created, that
it must be backed up, be unalterable etc.
<The records "profession" has ot defined metadata as being a required
component to ensure record integrity>
It has. The 2002 PRO (UK) section 2, Metadata, makes many statements like
this: page 2, section 1: "when properly implemented, records management
metadata does this (prevents alteration and proves where alteration has
occurred) by:
...establishing the provenance of the record....
....showing whether the record's integrity is intact....
.....demonstrating that the links between documents, held separately but
combining to make up a record, are present".
So in a project like J.'s I would preserve the metadata records from every
system the record has ever existed in permanently. The ideal would be
metadata harvesting ability between systems that capture the old metadata
into the new system - and treat that old metadata record with same controls
as audit metadata. Also the event of exporting records from one system to
another should ideally be captured as system-generated audit trail metadata.
What would any of you do?
Maureen Cusack, B.A.H., T.E.S.L., M.I.St.
http://www.maureencusack.net
List archives at http://lists.ufl.edu/archives/recmgmt-l.html
Contact [log in to unmask] for assistance
|