!071394 Electronic records as patent evidence: CFR 37, et.al. A few days ago, I posted the question of the validity of electronic notes and records in patenting procedures, where people use computers as electronic notebooks to record their activities while inventing a device. The general answer which I received was that it is doubtful that electronically stored notes and records can be used in patent proceedings, since such information is much more difficult to authenticate with regards to the original date of entry, or that the information is unadulterated. People's advice (some of which I mention below) was to always print out such electronically stored information as it is entered, and then treat the printed records as you would anything other document (notarized, etc). Before I summarize some of the comments I received, it would help to cite some patent law (which I should have looked at in the first place). The issue of electronic notes and records most likely occurs during interference proceedings, for which the CFR has a section: Code of Federal Regulations Volume 37 - CFR 37 1.671(b) Subpart E - Interference 1.671 Evidence must comply with rules 1.671(b) Except as otherwise provided in this subpart, the Federal Rules of Evidence shall apply to interference proceedings. Those portions of the Federal Rules of Evidence relating to criminal actions, juries, and other matters not relevant to interferences shall not apply. Now I haven't see the Federal Rules of Evidence, but I venture that in these rules there is something that disallows for the purpose of establishing a priority in time, the use of electronically stored information. (If anyone could check, I would appreciate it). So for interference purposes (and I suspect other patent activities), the Federal Rules of Evidence apply. At some point in future, when digital signature standards are approved, I suspect properly prepared electronic documents will be admissible (I think ATT had some trial escrow scheme where code words were published in the New York Times). Some of the comments: There is no set standard on this question, although printing out and notarizing or witnessing the printout is the best bet. The rule basically is that the data has to be intrinsically reliable both as to what was there and when it was put there. The problem is that the data and dates on the electronic files can so easily (and undetectably) be modified simply by resetting the date on your computer, then modifying and saving the file. FYI, for the last two years a consortium of about 19 companies has been looking into how to handle electronic records to ensure legal admissibility, without having to print them out (which can take an enormous amount of paper in some cases). Brad Huther, Associate Commissioner of the PTO, has set up a "matrix" group including several other agencies (AIIM, NIST, FDA) which will be meeting towards the end of August with the consortium to discuss their proposals. This eventually may lead to the adoption of standards by the agencies, especially NIST (National Institute of Standards & Technology, the official government standards agency). The consortium is being led by Dr. Kewal K. Likhyani. His address, phone and fax are: E.I. Du Pont de Nemours & Company, Inc. Fibers Chestnut Run Plaza - Laurel Run Wilmington, DE 19880-0705 Phone: 302-999-3845 Fax: 302-999-2805 *********************** Excellent point. I have taught several classes at the University of Texas on product promotion and that issue came up. The experience of several members was that the heavy, spine bound book in which the pages cannot be remove and having EACH major page notorized was the BEST way to document first-to-invent cases. Several inventors who have software that "achieve" the same purpose as the written log-book had a much harder time proving the credibility of the source since software entries are much easier to forge. This is one area where technology is not neccessarily better. We had discussed scanning in the inventors notebook into a desktop publishing program to create an electronic version but just an idea. *********************** Anything you can do to minimize the likelihood that someone can later argue that the records were made up only after the fact, will help. For example, your opponent will know that file "last modified" dates are easily changed after the fact in electronic records. So don't rely on that exclusively. A date on a printout is also easily modifyable after the fact with today's photocopiers, and also easily back-dated at the time it is printed. So don't rely exclusively on a date on a hard copy either. A notarization is a good idea, since the notarization date is put there by a third party notary who risks his license if he lies. He also keeps an independent record of the notarization, so if your opponent later says you modified the notarization date after the fact using a photocopier, you will be able to get the notary's own record to prove that you didn't. But if you're considering filing a patent application in the future, the best way to establish a date of invention is by using the PTO's Disclosure Document program. You send in your description, and they stamp it with a receive date and put it in a safe for up to two years until you file your patent application. There are some traps if you don't do it correctly, so read the MPEP carefully. [NOTE: the PTO's Disclosure Document program is a PTO service where you can send them any printed materials, and for $10 they will hold the records for two years, and testify that they received your materials on a specific date. Within two years, you have to file a patent citing this document, or the PTO destroys the records. By itself, the Disclosure Document offers no protection. Check Pressman's book for a better explanation of this service.] *********************** Boy what a great question! I don't have any idea and the Patent lawyers here at LLNL haven't even asked the question. An example for your thoughts: Invent some new well logging gear and make a disclosure in the usual way (paper). The disclosure references testing done to verify some theory used to design the preferred embodiment. All theory and the test are in the computer except for some handwritten notes used as a backup. This type of data which only exists in the computer (on the order of 100 Mbytes) has often been important in patent cases in the past. Even printing it out is economically infeasible not to mention silly. How would you -- 1. Establish the date of acquisition? Give copies to third party? 2. Authenticate or sign the data? 3. Assure that the data will survive the 20 or 30 years that may be required for patent litigation? Patent types here haven't even thought of this yet but I would expect that future patent cases may hinge on just these types of considerations. Current patent practice may need to think ahead on these issues to make defendable patents. *********************** It's my understanding (I don't have citations) that the PTO and courts have generally not accepted electronic records. The reason, however, is not that electronic records are _per se_ unacceptable, but that applicants are generally unable to adequately authenticate the date and content of their electronic records. In other words, the PTO and courts don't require records to be on any particular media (paper, electronic, magnetic, optical, etc.), but they do require records to be in a form that permits independent corroboration. Paper makes this relatively easy (if handled correctly), so the PTO and courts readily accept paper records. Non-paper media is easier to falsify, so the PTO and courts require greater proof that the records are legitimate. Therefore, patentees wishing to use electronic records must adopt methods that provide this proof. For example (to borrow from Johnny Carson's old character, "Karnac the Magnificent"): seal the electronic records in a mayonnaise jar kept on Funk & Wagner's porch. In other words, patentees do not have to use paper media if they can prove that their records are, nevertheless, immutable. Greg Aharonian Internet Patent News Service