March 1, 2005

It is inevitable that the electronic filing of documents in the civil courts will become the predominant mode of filing within a few years. The reason is that, for the majority of users (but not all), the advantages of electronic filing, as compared with paper filing, outweigh the disadvantages.

As with any IT initiative, the disadvantages which the electronic system potentially introduces must be recognised, designed out of the system where possible, or absorbed by changes in the business method. For example, I suggest below that e-mail, the simplest e-filing solution, has serious drawbacks which means that it will never be suitable for electronic filing.

Effective dialogue between users and those in government who may, in due course, become responsible for specifying and designing an e-filing system is fundamental. Members of SCL are uniquely placed to assist in that process. This article is intended to stimulate debate in the Society, and provide some ideas to move e-filing forward.


The Department of Constitutional Affairs has allocated no money to implement electronic filing of documents in the civil courts. It is frustrating because our criminal counterparts are given lots of money to spend on exciting IT projects: information screens in court centres, and wireless access for jurors for example. But they have also invested wisely in secure e-mail which permits the transfer of confidential documents between lawyers and government authorities.

Some money has been found by the DCA for the implementation of a pilot “electronic case file” in the Commercial Court. This project, which involves configuration of an existing US software product, is essentially a document management application with some diary management in there as well. Incoming paper documents are scanned by the court staff. E-filing forms no part of the project.[1]

The pilot is a very welcome step towards modernising the civil courts through IT. If the project is well received by court staff, and if it is within budget, perhaps the civil courts will be rewarded with funding for other projects, including potentially e-filing. However I do not believe that we, as interested members of the public, should wait until then. We need to be in a position to have developed our own ideas on e-filing to assist government even before the success of the Commercial Court pilot is known.


I have suggested that e-mail is inappropriate for e-filing. Why? The reason is that for some purposes, for example in cases involving a limitation period, a reliable time stamp on a document is essential. If a claimant sent a claim form by e-mail to the court, the claimant could not be sure that the email would be received at the court’s computer before expiry of the limitation period. Even if you changed the court’s rules so that the time at which the e-mail was sent, as recorded on the e-mail by the originating computer’s clock, was deemed the date and time of issue of the claim form, the originating computer’s clock may be incorrect.

If in doubt, software designers should always learn from the manual process which has usually developed over many years. The manual process in this instance, that of stamping the claim form at the court’s offices, is completely reliable – the claimant knows the business hours of the office; if he turns up at the office between those hours, his document will be stamped. If he fails to do so, that’s his problem. It is important to bear in mind that with the current manual process, all of the risk of missing a time limit rests with the user. With e-filing on the other hand, there is potential “uncertainty in the post”: not the court’s fault and not the claimant’s fault either.

Of course, with conventional post there is uncertainty, and CPR 7PD5.4 says that parties proposing to start a claim which is approaching the expiry of a limitation period should recognise the potential importance of establishing the date the claim form was received by the court and should themselves make arrangements to record the date (ie use recorded delivery).[2] It would be inconvenient to establish the date of receipt of an e-mail at court. Perhaps forensic evidence of the state of the court’s computer would be needed. Whether this is so or not, a claimant whose limitation date is approaching cannot rely on e-mail any more than post to ensure that a document is filed in time. He must go to the court office.

It would be possible to have a rule of court which said that if an email had been sent but not received by the court it would be deemed to be received within say, two hours. Then a claimant could rely on the rule, send the document at two hours to midnight, and no-one would need to have to go to the court office. However such a rule would again mean establishing the date and time at which an email was sent. Unlike registered post, there is no independent authority which will verify (a) that an e-mail has been sent and (b) when it was sent. Further the unscrupulous claimant would not resist the temptation to adjust his computer’s clock to obtain a favourable time. E-mail is just not up to the job of e-filing.


Once one takes the position that the user’s computer clock is inherently unreliable evidence as to the time at which a communication is sent, one is thrown back to relying on, and proving, the time of receipt at court. This is easily achieved by transporting the document over a real-time connection, analogous to a fax connection, between the claimant and the court. In this way, everyone knows that the time of receipt is the same as the time of transmission.

This method raises new problems. Suppose the court’s server is unavailable? Simple: rules of court could deem filings made on the day after the server was working again as being issued on the first day the server was down, the court having issued an official circular that the computer was down for that period (like the rules extending postal times where the court certifies that there was a postal strike).

What should the situation be if a transmission starts at 5 minutes to midnight but completes at 5 minutes past midnight. Simple: rules should deem the start of the transmission, as recorded by the court’s computer, to be the relevant time, even if the transmission could be expected to take much longer.

Suppose the court’s server is available, but for some unknown reason the transmission fails. The Patent Office, which has implemented e-filing of patent applications in circumstances where again the time stamp is critical, has suggested referring back to the sender’s records of date and time in these circumstances; but this is open to the objection discussed above that it does not take into account the vagaries of the sender’s computer clock. I believe that a combination of (a) evidence from the court’s computer as to the time of commencement of the failed transmission coupled with (b) a general discretion, on the appropriate evidence from the claimant, to deem the date of issue as being the date of an attempted but failed transmission, is the solution here.

What is important is that rules should exist such that the claimant should be able to rely upon a commencement of the transmission, for example shortly before midnight, as complying with the Limitation Act, even if the transmission fails. There is still scope for fraud (by the claimant who, realising that midnight is fast-approaching, starts, and then aborts, a fictitious transmission to give the impression that a claim form was attempted to be filed) but less scope for fraud than with e-mail.


Despite appearances, paper systems are not strong in terms of ensuring that documents which are filed truly originate from the person from whom they purport to originate. The incidence of fraudulent lodging of paper documents is so low that the courts have never seen the need to authenticate the origin of paper documents, eg by requiring filing to be conducted by authorised representatives only, whose identity can be verified by a reputable agency such as a professional association.

I do not believe that any more authentication is needed when e-filing than is used with paper documents. It is true that, where there is a problem with “identity theft”, e-filing can improve upon paper systems. Companies House has responded to the phenomenon of company hijacking (in which fraudsters send in documents to change details of companies prior to using them as vehicles for purchasing goods which are not paid for) by adopting e-filing in order to benefit from its authentication capability.[3]

The Court Service’s existing Money Claims Online[4] is interesting. There is some authentication, even though the claims are of low value. Rightly, a decision has been made to keep the authentication weak, in the form of merely requiring a valid e-mail address; stronger authentication, ie based on limiting the class of permitted users to lawyers only, prevents e-filing being used by members of the public. The risk of identity theft when filing court documents does not justify this restriction, in my view.


Coupled with the question of authentication is the question of security. The threat is of the malicious or misguided litigant in person or hacker. The threat takes two forms: (a) an attack on the system so as to introduce a virus or to bring down the service; (b) an allegation of tampering with the electronic copy of the document held at court.

As to the first of these, this must be met with the use of a reliable server infrastructure, using operating system and other software whose security strengths and weaknesses are well understood, and which can be locked down.

In relation to tampering, an e-filing system must return to the user, either by e-mail or immediately as the document is filed, a receipt containing a “fingerprint” of the filed document, so that any tampering can be disproved. Copy fingerprints would be stored by the court with the document. This is well-known technology.

Other necessary components

An e-filing system on its own is of limited value. Until an electronic court file is in place, electronic documents will have to be printed out and placed in the manual court file. This is not an argument for waiting until an electronic court file is built before embarking on e-filing – in fact a computer system’s ability to integrate with a manual system with minimal disruption to the business process is a sign of good design – it just means that you have to make sure there are industrial quality printers available to the court staff to support the project.

For many documents filed at court, a fee is payable. Any e-filing system must be able to integrate with computer systems, as yet unspecified, handling invoicing and other financial matters.

I believe that IT in the courts should be built from small free-standing network components mounted on discrete micro-computers[5] in which each function is decoupled from other functions from a programming point of view, but communicates with other functions using open standards over a network. This will allow competition from software vendors at every point of extension, will simplify and therefore improve testing at the component level, and enable resilience. These factors reduce the risk to the courts as purchaser[6] and also ease the pressure on funding, since components can be plugged in as and when there is money to pay for them.[7]

Most importantly (because technology only works with the support of the people who have to use it), this model enables court staff (a) to be trained on manageably sized sub-components rather than learning to drive a juggernaut; and (b) to be participants in a small, but quick, success, rather than promising a large success over an ever-receding horizon, which is what the mainframe model usually provides.

Once the court has built up a number of self-standing network components for such things as e-filing, fees, diary and document management, the connections (or “choreography” to use the jargon) between these can gradually be increased when the courts, and the court staff, are ready to phase out manual systems. There is no “big bang” to worry about.

Many candidate implementations of this pluggable network architecture exist today. One of them, the SOAP Web services standard, which is administered by the respected W3C consortium, defines how two components in a computer system should communicate with each other over a network and looks to me to be a very strong candidate technology. Another is Sun Microsystems’s Java/RMI/JINI technology, but this narrows the universe of potential software vendors to Java specialists.

As to the open standard for containing data within the system, LegalXML is the obvious choice.[8] The LegalXML consortium has done most of the expensive business analysis work to arrive at a comprehensive data structure, and made it available for anyone to use. Vendors can build software more cheaply if the standard is used. LegalXML, like any XML, can be transported between SOAP Web services.

Suggested implementation of electronic filing at Court

Pictures speak a thousand words, so I have prepared an alpha implementation of a candidate system for electronic filing at court. It is a SOAP web service at http://lawdata.xbundle.com/efiling/SubmitFiling.[9]

The XML messages themselves are in the LegalXML standard, with the addition that the receipt includes a standard “MD5 with RSA” fingerprint for the document integrity. A typical LegalXML receipt generated by the system is shown in Figure 1.

Any valid LegalXML Court filing within the SOAP message body will be accepted by the service. The files are deposited on the server within a directory corresponding to the case number – no attempt is made to implement any case management service.[10]

This alpha system would therefore be suitable for a small pilot project involving a subset of users for example solicitors or barristers wishing to file skeleton arguments at a specialist subdivision of the High Court, so that the court and practitioners can get a feel for how electronic filing might work in practice. For the pilot, the documents arriving at the server would simply be fetched by secure FTP, printed out by the judge’s clerk and given to the judge, the files then being moved by secure FTP to an archive directory.

The service is stream based. This means that the LegalXML message is written onto the hard drive as it is streamed off the sender’s machine, and the receipt is simultaneously streamed back on to the sender’s machine. In this way the receipt is received by the sender simultaneously with the reception of the document by the receiver; and it also overcomes the need to read the whole document into the server’s memory, which would otherwise effectively prevent very large documents being filed.[11]

Since the system is based on SOAP and LegalXML, anyone is entitled to implement a SOAP client to post LegalXML filings into the service. I invite potential software vendors to do so. I hope that this will demonstrate graphically how open standards effectively create a market in sub-components which eventually will drive down prices for the purchaser (ie the Court Service). I have developed such a client at http://lawdata.xbundle.com/cocoon/efiling/EfilingClient and readers are invited to try it out. Please do not send confidential documents.


Electronic filing of documents as the principle method of getting documents to court is inevitable. There is presently no government money to fund any of this. It is up to individuals to make the running. I hope this paper will encourage SCL members to engage in the debate and support any initiatives which anyone is prepared to undertake in this field.

Robert Onslow is a barrister at 8 New Square, Lincolns Inn, practising in the field of intellectual property and information technology law. He is the author of software in the legal field including XBundle (2004) (xbundle.com) and XFiling (2005) (lawdata.xbundle.com/cocoon/efiling/EfilingClient).

[1] It is important to bear in mind that e-filing and electronic case files are different things: e-filing is concerned with transporting documents from users into court; electronic case files is about replacing the current manual files used by the courts to manage cases, and raises wholly different, and difficult, issues. A good e-filing system should be designed to integrate both with the manual court files currently used by the courts, and with such electronic case files as may, in the future, be used by the courts.

[2] Perhaps they would do rather better to consider whether the document will arrive in time or at all.

[3] http://www.companieshouse.gov.uk/infoAndGuide/proof.shtml

[4] www.moneyclaim.gov.uk

[5] Eg one e-filing “cluster” for each division of the High Court, one host for each specialist court, and one host for each county court – this simplified administration.

[6] The courts do not want to repeat the well-known problems experienced by government procured behemoth computer systems modelled on out of date mainframe technology.

[7] Software using open standards is cheaper to build than proprietary software, because (a) the developer can use off the shelf sub-components; and (b) there is downward pressure on prices caused by proper competition. However open software is neither free software, nor is it shoddy software: in fact open source subcomponents have generally had more opportunity for testing than proprietary components.

[8] There may have to be some extensions to the standard to cater for the UK courts’ requirements.

[9] Readers should not browse directly onto this URL – it is a POST-ing URL. Browse onto the Efiling Client URL set out below.

[10] If this service were to be used, a software vendor should be invited to provide a second SOAP Web service on the e-filing server which would allow case management software to call off the documents from the e-filing server into the case management system. This is a small and cheap job, and illustrates how the use of Web services allows the system to be gradually scaled up by the addition of cheap micro-components.

[11] Image documents could be 50MB or even 100MB in size. One day, entire sets of electronic bundles will be e-filed in time for a court hearing, giving file sizes in the region of 1GB.