Protection of Information Under the EU Sui Generis Right for Databases: Some Implications of the Decision in British Horseracing Board Limited and Others v William Hill Organization Limited
On 9 February 2001 Mr Justice Laddie gave judgment in this, the first reported decision on the extent of the sui generis right created by the European Union and intended to protect databases throughout the European Economic Area. The decision is subject to a pending appeal, and is based exclusively on the EU Database Directive which defined the new right.
The British Horseracing Board Limited ("BHB") as governing authority for the British racing industry maintains at substantial cost a large database of information related to racing, from which BHB earns substantial fee revenue from third parties for the use of information contained in the database. Much of this information is made publicly available with BHB's consent through newspapers, Internet websites and by other means.
William Hill Organization Limited ("WHO") runs licensed betting offices and provides a telephone betting service for which purposes it was using publicly available information derived directly or indirectly from BHB's database: in particular a declarations feed, accurately detailing runners and other last minute details for each significant race, which was made available to subscribers to the BHB database. Information from this source was used by WHO at its licensed betting offices and for its telephone betting service without objection from BHB, which received, directly or indirectly, financial compensation for such use.
In 1996 WHO started to provide an Internet betting service, for which it used similar information also derived, directly or indirectly, from BHB's database. This latter new use was without BHB's licence or other express consent. BHB objected, but WHO claimed that the information used comprised facts which were publicly available and which, accordingly, WHO was free to use without BHB's licence.
Laddie J held that database right, as established under the Copyright and Rights in Databases Regulations 1997 transposing into UK law mandatory provisions of EU Directive 96/9/EC on the legal protection of databases, prohibited the unauthorised extraction and/or re-utilisation of information contained in a database qualifying for protection under the new EU sui generis right. The BHB database qualified for this protection, and WHO's unlicensed use of information directly or indirectly derived from BHB's database infringed BHB's database right notwithstanding that similar information from the same source was publicly available.
The facts, as disclosed by the approved and published transcript of Laddie J's judgment, show that BHB, the first named claimant, was owned and controlled by its members which included the Jockey Club and other parties concerned with regulating the British horseracing industry. These regulatory functions were now vested in BHB which compiled data relating to horseracing and annually produced a fixture list in July of each year for the next succeeding year. This process culminated in weekly advertisements for races some 7,800 of which were run each year. The number of entries for each race varied, and changes might be made in runners entered for any race before the race was run, for example if a horse was withdrawn. Shortly before each race, trainers were required to make declarations of their horses which were to run in each race. In 2000, there were 80,000 such declarations.
BHB undertook its functions partly by means of a database (the "BHB database") which was in a state of constant updating. The BHB database contained information about horses, owners, trainers, jockeys, colours, fixtures and pre-race information (date, time and place, distance, eligibility etc). Shortly before each race, related pre-race information was finally updated to show declared runners, riders, and other details setting out the latest position about that race and the horses which were to compete in it. This final updating was important information on which bookmakers and their customers relied in taking and making bets.
The cost of creating and updating the BHB database accounted for about 25% of BHB's total annual expenditure of £15m. BHB obtained its income principally from fees, registration and licences, including fees charged to third parties for the use of information contained in the BHB database. These latter fees currently earned just over £1m annually, meeting in the region of 25% of BHB's costs of maintaining the database.
The information contained in the BHB database was of wide interest not only to participating elements of the horseracing industry, but also to newspapers and other media and to members of the public.
Selected parts of the database's information were supplied by BHB to interested parties, including book-makers, and in particular to Racing Pages Limited which was controlled and owned by Weatherbys (the third named claimant) and the Press Association. Racing Pages Limited, on behalf of BHB, provided a selection of this information in electronic form to its own subscribers, including a declarations feed normally made available on the day before each race and containing up-to-the-minute lists of declared runners, jockeys, race times and other information. Data were also supplied by BHB to Satellite Information Services Limited ("SIS") which was allowed to use the data for certain purposes including onward transmission to, and for use by, SIS's own subscribers. This onward transmission by SIS was known as a Raw Data Feed ("RDF").
WHO was an important provider of off-course book-making services, including the running of licensed booking offices ("LBOs") and the provision of a telephone betting service. SIS provided live audio and video coverage of horseracing events, which WHO made available at its LBOs with additional pre-event information for each race, in the form of a text service on television screens which were regularly updated. The same information was also available at each LBO in the form of printed pages from the betting office edition of the Racing Post, which pages were posted on LBO walls. The latter information was more comprehensive than that available on LBO screens and included details of owners, trainers, colours and other information. Copies of the newsstand edition of Racing Post were also available at each LBO.
A great deal of the information used by WHO at its LBOs came directly or indirectly from the BHB database. BHB took no objection to such use which BHB accepted to be with its express or implied consent and for which BHB received, directly or indirectly, financial compensation. The most critical information used was that derived from the declarations feed, which WHO used for its telephone betting service. It was the immediacy, up-to-dateness and accuracy of this information which was so important to the placing and acceptance of bets before any given race.
Although BHB had raised no objection to WHO's use of information derived from the BHB database when that use was restricted to use at LBOs and for the telephone betting service, BHB objected when the same information began to be used by WHO, without express licence from or further payment to BHB, for the purpose of a new on-line betting service which WHO started to provide by means of an Internet website. By March 2000 members of the public were able to access two WHO Internet sites and to place bets electronically. Relevant data for all races in the calendar became available at these sites and were derived by WHO from the RDF supplied to WHO by SIS. By the time these data had been published by WHO on its Internet site the same information was available from sources other than SIS.
WHO did not dispute that the information displayed on its Internet sites was also to be found on the RDF feed received by WHO from SIS, nor did WHO deny that the same information was supplied to SIS from data stored on BHB's database. It was not in dispute that SIS had no right to sub-licence WHO to use any of BHB's data on WHO's Internet site, nor did WHO claim that SIS had purported to grant any such licence. However, WHO did not admit that the information used by WHO on its Internet sites originated from the information supplied by SIS and derived from the BHB database, even though the information used by WHO was identical to it.
BHB claimed that it owned the database right in the BHB database, and that WHO was making unlicensed use of those data in its Internet betting business: BHB also claimed that what WHO took from the SIS RDF was derived from the BHB database. These actions, said BHB, infringed BHB's database right in two ways:
- each day's use by WHO of data taken from the BHB database was an extraction and/or re-utilisation of a substantial part of the contents of the BHB database, contrary to Article 7(1) of the Database Directive; and
- even if the individual extracts taken were not substantial parts of the BHB database, WHO's actions amounted to repeated and systematic extraction and/or re-utilisation of insubstantial parts of the database which conflicted with BHB's normal exploitation of it and which unreasonably prejudiced BHB's legitimate interests, contrary to Article 7(5).
WHO's defences to these claims were:
- BHB had failed to prove that the information used by WHO had been derived from the BHB database. The fact that the information used by WHO was identical in content to information contained within the BHB database did not prove that the information so used had been derived from BHB: SIS may have derived the information from another source before passing the information on to WHO;
- such information as was used by WHO was not a part of the BHB database, or if it was a part it was not a substantial part;
- the use made by WHO of the information was not either extraction or re-utilisation from a database, in the terms of Article 7(1) of the Database Directive;
- even if WHO's use of the relevant information was extraction or re-utilisation for the purposes of the Database Directive (which was denied) the consequence of BHB's daily updating of its database was to create not one database but many successive databases, so that WHO's daily extractions and re-utilisations were not repeated and systematic and from one database contrary to Article 7(5) of the Database Directive but were successive extractions and re-utilisations of insubstantial parts from a series of different, but interrelated, databases;
- WHO's actions in taking insubstantial parts of one or more databases did not conflict with BHB's normal exploitation of its database or databases and did not unreasonably prejudice BHB's interests as a maker of databases as restricted by Article 7(5) of the Database Directive; and finally
- facts as such may not be protected by database right. If WHO was debarred from using information taken from the BHB database in the form in which that information appeared in it, WHO would modify its website so as not to use the relevant information in a form in which it appeared in the BHB database, and having modified the information WHO would then be free to continue to use it.
As to whether the information used by WHO was a part, and a substantial part, of the BHB database, WHO distinguished between data or information within a database and the characteristics which gave rise to protection under database right. WHO claimed that database right did not protect information as such, and that BHB could not use any database right which it may own to prevent WHO, or anyone else, from making use of facts which appeared within the BHB database. Recital 46 of the Database Directive provides that the existence of a right to prevent unauthorised extraction and/or re-utilisation of works, data or materials from a database should not give rise to the creation of a new right in the works, data or materials themselves. This meant, said WHO, that since no right was to be created in the works, data or other materials contained in the database the protected 'database-ness' of a database must lie in the fact that the independent works, data or other materials in the database were arranged in a systematic or methodical way, and were individually accessible. WHO's argument continued that any acts which did not make use of the arrangement of the contents of the database, or take advantage of the way in which the database maker had rendered the contents of the database individually accessible, could not infringe database right. It followed that that which could not attract database right except by reason of its arrangement and/or individual accessibility would, when robbed of that arrangement and accessibility, not be subject to database right, and that therefore the courts could not treat its unlicensed copying or use to be an infringement.
As to WHO's argument on derivation of the information used at WHO's Internet sites, Laddie J was satisfied on the evidence that all the information in the racing calendar used by WHO was ultimately derived from the BHB database, and in particular that vital information found in the RDF and used by WHO could not have come from any alternative source.
As to WHO's distinction between unprotectable data or information within a database and the characteristics giving rise to protection of a database, Laddie J did not accept the argument that only 'database-ness' was protectable. He distinguished between the requirement of form, which must be present before a database will be recognised as existing, and the features of content or investment which were protected once a database had been held to exist.
As to whether the information used by WHO was a substantial part of the BHB database, Laddie J held that this issue was to be assessed primarily by comparing what had been taken or used with what was in BHB's database, but that the importance of the information to the alleged infringer was also relevant. Both the quantity and the quality of what had been taken were to be assessed together. The data taken by WHO relating to particular races were the most up-to-date core information available within the BHB database. WHO was relying on, and taking advantage of, the completeness and accuracy of the information taken from the RDF, which was a product of BHB's investment in obtaining and verifying those data: it was thus a substantial part of the BHB database's contents.
Laddie J found that the use made by WHO of information derived from BHB s database was both extraction and re-utilisation. WHO had argued that extraction means an act of 'taking away', and that once a thing had been extracted or taken away from somewhere it could not be taken away again: therefore, it was only the first removal of data which amounted to extraction. When, for example, Racing Post under licence made information available to the public anyone was thereafter free to copy that information from the Racing Post without BHB's licence and without extracting information from BHB's database.
Laddie J found against these submissions. The Directive did not require that extraction should be direct rather than indirect, nor did the Directive's definition of extraction involve the concept of taking away. All that was required was that a substantial part of a database's contents be transferred to another medium. It mattered not that, when a database was copied onto a new medium, the original database was left behind. However, the requirement of transfer to 'another medium' may have the consequence that a hacker who looks at data and memorises but does not copy its material may not be guilty of extraction.
As to the submission that BHB had not one but a series of related databases, so that repeated and systematic extracted would not have been from a single database, Laddie J found that BHB had a single database which was in a constant state of refinement. Its contents changed from time to time and any attempt to split it into a series of discrete databases, besides being impossible to do, would not reflect reality.
As to WHO's expressed intention to re-format the information on its websites so to avoid the 'database-ness' of the BHB database, Laddie J concluded that what WHO had in mind involved the manipulation of information the unlicensed taking and use of which was protected by database right. If a database happened to be written in English, he said, an unlicensed third party who displayed a substantial part of it would not avoid infringement by doing so in French, in German or in Chinese ideograms, nor would he avoid infringement if he translated information in denary code to its binary equivalent. As long as substantially the same information as had been derived from the BHB database was made available on the WHO website, the same acts of extraction and re-utilisation would have taken place.
The Database Directive, with its preparatory materials, is the sole source of the sui generis right and so of database right.
The central issue was, and remains, the correct interpretation of the Database Directive and in particular its Article 7. This Article, headed "Object of protection", is potentially ambiguous both as to its intention and as to its effect. Recital 40 of the Directive states that:
- "whereas the object of this sui generis right is to ensure protection of any investment in obtaining, verifying or presenting the contents of a database for the limited duration of the right; whereas such investment may consist in the deployment of financial resources and/or the expending of time, effort and energy"; and
- Recital 41 states:
"whereas the objective of the sui generis right is to give the maker of a database the option of preventing the unauthorised extraction and/or re-utilisation of all or a substantial part of the contents of that database; ...
The terms "object" and "objective" as used in these Recitals are confusing. Article 7 is headed "Object of protection" but it is unclear what is intended by the Directive to be protected. Laddie J has construed the object of protection, in the sense of what is to be protected, as the database maker's investment in a database, such protection to be achieved by restricting extraction and/or re-utilisation of the whole or a substantial part of the database's contents. He might have construed the reference to substantial investment as a requirement of or precondition for the provision of protection under the sui generis right, in the same way as intellectual creativity is required as a precondition for the protection of a database by copyright under Article 3(1) of the Database Directive. This would have left the sui generis protection to bite on the whole or any substantial part of the database, so that restriction of extraction or re-utilisation would have protected the database itself with the consequent effect of protecting its maker's investment. Had Laddie J taken the latter interpretation it would have been more difficult to justify protecting information inherent in a substantial part of the BHB database, since to do so would have been to create a new right in data or other materials comprised within the database's contents, contrary to Recital 46 of the Database Directive. 'Materials' construed eiusdem generis with the preceding "works" and "data" is broad, and could be taken to include information.
On that latter construction, to protect information included in, or capable of being derived from, a protected database would have been to acknowledge the creation of a new right in materials comprising information: since information cannot be owned, such a right would certainly be newly created.
However, Laddie J took the view that the sui generis right's object of protection is the database maker's investment and that the protection of information derived from BHB's database was necessary in order to protect BHB's investment in obtaining, verifying and presenting the database's contents, and so was in accordance with Article 7(1) of the Directive.
There are other significant points to be observed:
- the analysis of BHB's database as a single entity in a constant state of amendment may be contrasted with databases which are updated on a batch or edition basis, as for example annually updated directories. There may be difficulty in determining whether a collection is a single organic database or a series of successive and related databases in the form of successive editions: some databases may be updated once an hour, once a month, or once a year. WHO's argument on repeated and systematic extraction could have applied if both extraction and updating had been infrequent so that there had been systematic extraction only from a series of related but distinct databases.
- Laddie J's conclusion that modification of information extracted from a protected database will not prevent infringement if extraction or re-utilisation falls within either Article 7(1) or Article 7(5) of the Database Directive heightens the significance of treating information derived from a database as protected. It follows from that conclusion that the deduction of meta-information (new information derived from, but not explicitly contained in, a database) from a database, provided always that the meta-information is transferred to another medium, may be classified as extraction or re-utilisation so widening the protection given by the sui generis right to investment in databases. Such widening may be entirely fair: if meta-information can only be cost-effectively obtained from an expensively compiled database, why should not the investing database maker be protected? But the effect of recognising such protection may be to monopolise publicly accessible sources of information, as well as information itself.
Conclusions and Implications
On the facts, and on the basis that the object of protection under the sui generis right is the database maker's investment, the immediate consequence of Laddie J's judgment in favour of BHB seems just. But is the principle too broad? Could it result in wider protection being made available for information than was intended by the Directive? What if the defendant had been an innocent user of data extracted by a third party, unaware of BHB's database? How far may information extracted without licence from a protected database be followed by the database's maker? What about split, deduced and mixed information?
Laddie J stressed the protection of investment, that the required investment must be in the obtaining, verification or presentation of the contents of a database, and that expenditure on making data would not qualify as investment for the purpose of qualifying a database for protection by the sui generis right. It may be difficult to draw a line between investment in the making of information, and investment in obtaining it. Suppose, for example, that substantial investment is made in the production of computer-generated information: is the resulting information made or obtained? Will the investment qualify for protection under the sui generis right? Must the investment have been made in obtaining, verifying or presenting that part of a database which has been the subject of extraction or re-utilisation, and if not will protection extend to information the obtaining, verification or presentation of which required little or no investment but which forms a substantial part of a database which otherwise required substantial investment?
Increasingly, databases contain computer-generated data, and databases which are computer-generated will be debarred by the Database Directive from protection by copyright if they are not their author's own intellectual creation. For such databases, when they are made publicly accessible, the sui generis right remains the only readily available form of protection. That protection is enhanced in importance by Laddie J's decision that the sui generis right protects information directly or indirectly obtained from a sui generis right protected database, notwithstanding that such information may be generally known or otherwise publicly accessible. May those who invest in the obtaining of information which they wish both to make publicly available, but also to continue to control, use the sui generis right to secure such continuing control by ensuring that substantial investment is made in the obtaining, verification and presentation of such information as a substantial part of the contents of a protected database? This may put within reach the Holy Grail, for some, of protecting information as such even after that information has become publicly available and generally known, and so incapable of protection under the law of confidence.
Using Database Right to Protect Information
Those who provide information to others, and those who use information provided by others, may wish to review their existing business practices and to take the following points into account.
For information providers, database right potentially offers a new form of protection: it can give them the ability to restrict the re-use of their information by those to whom it is made available, including those who gain access to the information without having had any contractual relationship with the original information provider. Database right potentially extends to protect underlying information as such, including information derived from published information even where the derivation is indirect.
For database right to be available the information must be comprised in or derived from a substantial part of a database, defined as a collection of independent works, data or other materials arranged in a systematic or methodical way and individually accessible by electronic or other means. The right is only available to protect databases which fall within this definition, though such databases need not necessarily be electronic. There must however have been substantial investment in either the obtaining, the verification or the presentation of the contents of such a database to qualify the database for protection under the new database right.
Potential beneficiaries of this new kind of protection include publishers of directories and reference materials, compilers of statistics and information, weather forecasters, travel agents, cartographers and business and financial commentators. Where previously information purveyors of this kind have had to rely on contract, confidence or copyright for the protection of their materials, they may now be able to claim database right as an entirely new form of protection.
Conversely, those who take and re-publish or otherwise re-use and make available information derived from elsewhere may find themselves at risk of infringing third party database rights in the information which they obtain and re-use. This risk is increased if the information in question originally derived from a source of which the current user is unaware. Information re-users will need to discover the source of their re-used information and to make sure either that they have permission for its re-use or that no permission is needed from any third party database rightholder. When preparing or compiling information for publication, its provenance should be established and evidence of substantial investment in its obtaining, verification or presentation should be preserved in order to avoid the risk of infringing third party rights and to qualify the re-use of information in question for its own protection.
If a right to re-use publicly available information is challenged, consideration should be given to the possibility of the challenge being an abuse of competition law. To use database right in order to restrict the re-use of information which is publicly available but which is derived from a single source could breach European competition rules, and licences to re-use information may be compulsorily available.
To prevent inadvertent re-use of information in breach of database right, publishers should mark their information with a note that database right is reserved, with the year of first compilation. Database right lasts for an initial fifteen year term, but may be extended for further similar terms if substantial investment is made in the verification or re-presentation of the contents of a database.