I like this article here about a judge’s extensive evidentiary analysis of the admissibility of electronically generated data, in particular the fact that the judge found that certain types of computer generated data are not subject in any manner, shape or form to the rule against hearsay. As the article sums up the judge’s point: 

Perhaps the more important — and interesting — aspect of [U.S. Magistrate Judge] Grimm’s opinion is his conclusion that information he describes as "electronically generated" is completely outside the hearsay rule. The hearsay rule is among the best known and important of the gate-keeping rules to screen evidence before it reaches the jury. Under the hearsay rule, a statement made outside of court may not be offered into evidence for its truth unless the statement falls within one or more specific exceptions (for example, when the statements are contained in business records, made by employees or agents, or made against the speaker’s interests). Hearsay evidence is excluded at trial because there is no opportunity to cross-examine its creator to determine how reliable the evidence is. By electronically generated information, Grimm means information a computer creates itself. Electronically generated information can take a variety of forms. Grimm cites as an example the report a fax machine prints whenever a fax is sent. Another example (but which he does not cite) is "metadata." Metadata is information "created" by the computer that records (often without the user’s knowledge and often without the user ever seeing it) what has happened to a particular document, such as who created the document, when it was created, who viewed it and who changed it.

A neat analysis of some of the ever evolving issues arising from the expansion of discovery into computer data, but I can’t say I really buy it. Even though the data and/or the information printed out is not directly inputted by a person, the computer system itself obviously didn’t create and store these types of information on its own or create it from some artificial intelligence of its own. The information is there, and stored and formatted in a particular manner, only because of software and other input placed there by humans for the purpose of having the system create, store and generate the data in a particular manner. As such, the information discussed by Judge Grimm is at root the creation, by extension, of the individuals who did that, rendering the computer generated information a statement of some human or another. The information is therefore more accurately thought of as the out of court statements of an individual, and should only be admissible if an applicable exception to the hearsay rule applies – most likely, as a business record of the party in control of the computer network that generated the data or that created the stored information.

In any event, beyond this one quibble over the application of the hearsay rule to certain forms of computer generated data, the judge’s full opinion on these issues, in the case of Lorraine v. Markel American Insurance Company, is actually terrific and I recommend it to anyone interested in electronic discovery and the rules of evidence. Indeed, it is really a terrific treatise not just on admissibility of electronically stored data, but on the federal rules of evidence themselves. And as I have mentioned before in other posts on the subject of electronic discovery, the evolving law on these types of issues is central to both ERISA and insurance litigation, the primary subject matters of this blog, due to the fact that the administration of ERISA plans and insurance claims are almost always computer based.