The way of information dissemination has been, and will be continuously changed by the rapidly developing Internet technology and the ever newer modes of network operation. This brings impact and challenges to the existing system of law on Internet copyright issues and the court adjudication over related cases. Opinions vary regarding a few practical issues in court hearing of Internet copyright cases, but courts cannot wait for a conclusion before making their decisions . Now, we provide, as follows, our studies, understandings and doctrines in handling these issues, to ask for enlightenments from the wiser.

 

 

Among disputes over infringement of network dissemination right, the fundamental issue is whether an alleged infringing act is direct or contributory in disseminating copyrighted works on the Internet. The solution directly decides how to build a legal scheme of infringements under network dissemination right. In Poly Film Investment Co., Ltd. v. China Netcom Group Corporation Limited, Netcom was registered as an ICP in the ICP/ IP Address Information Filing and Management System of the Ministry of Industry and Information Technology (MIIT), but commissioned a third party to collect and upload works to its server for its users. Netcom argued that it only provided network services and this did not constitute infringement. It further defended itself by stating that the copyrighted works were uploaded by the third party and it only leased out a server and provided such a third party with the access to such a server. In Sanmianxiang Co., Ltd. v. Duowan Co., Ltd., the game website of Duowan contained a channel called “TXT Novels” where registered users might upload books in electronic formats. It also included such other channels as “Hot Categories”, “Usual Keywords”, “Hot Novels” and “Billboards”, and provided recommended download services. Moreover, it converted materials uploaded by users in other formats into the TXT format, and provided the converted materials for downloading by other users. The court ruled that Duowan was an ICP on the ground that it had edited the contents on its website by categorizing, creating downloading billboards and converting formats of such contents.

 

We argue that to infringe network dissemination right should mean that an infringer offers infringing works by wired or wireless means, so that the general public may receive such works at a time and place that they may choose individually. Here, (1) to “offer” means that the infringer provides infringing works on or through the Internet other than by any other means; (2) to “receive” means that any netizen may possibly receive or not receive such works; (3) most importantly, to “offer” is an equivalent of to “copy”, that is, the infringing works must be uploaded (i.e. copied) to the server of the infringer, and any other act vastly different from the copying act in nature is not deemed as an offering act; and (4) the server to which the infringing works are copied is open to the general public, so that the general public can access the server and obtain such works at any time and place which they may choose. Only on the above basis can a system of legal liability, including the liability on contributors or instigators, be established for infringement of network dissemination right.

 

In the Netcom case, the relationship between the third party which uploaded infringing works to Netcom’s server and Netcom which allowed the third party to upload infringing contents to its server for its users is similar to that between a client and its processer. Netcom and the third party, which worked together to offer infringing works for the general public, were joint transmitters of such infringing works. In the Duowan case, Duowan, which only provided a space to store information, but never uploaded (or copied) any infringing works to its server, was not a provider of infringing contents. The court erroneously ruled that Duowen’s categorizing, creating of downloading billboards or converting the format of the contents on its server was an act of uploading infringing works to its server on the Internet, for the purpose of an offering act under network dissemination right. Article 22 of the Ordinance for the Protection of Information Network Dissemination Right provides that a provider of information storage service which does not modify or otherwise change works that users provide may be exempted from infringement liability from the provision of such works. Although to convert the formats of electronic books that users uploaded is an act to change such works, this is only a required element to reject a safe harbor, but insufficient to establish an Internet service provider as an Internet contents provider.

 

Quite a few scholars have maintained that it is not necessary to create a distinction between direct and contributory infringement for the purpose of network dissemination right, and the act and liability of an infringer can be determined by generally relying on the four elements of infringement liability. We argue that the major reason for differentiating between a contents providing act and a service providing act, or between a direct infringement and a contributory infringement, is that the contents providing act, or the direct infringement, is an unauthorized copying act for the purpose of network dissemination. Generally speaking, an uploader is knowingly faulty, while an ISP (in a contributory infringement), being more neutral, may know or should have known, or never knows or should have never known, the infringement. In the latter case, a fault finding must be done on a case-by-case basis. Moreover, an ICP which provides infringing contents more probably contains a more obvious malice towards infringement. It should be penalized by law without mercy, so as to block unauthorized sources. An ISP, however, may have varying attitudes towards infringement. Its fault, if any, should be studied separately, in order to protect a good-faith ISP and promote the healthy growth of the Internet. Or, if the view prevailed that no differentiation between direct and contributory infringement would lead to higher efficiency and more severe crackdown on copyright infringement on the Internet, courts would go to an extreme to deal with the facts and acts of infringers in an oversimplified manner and without discrimination.

 

 

Deep linking displays all or part of the contents on a linked webpage through a hyperlink technology. Usually, users will not perceive deep linking, or notice the process of directing to the contents of the linked webpage. The address bar of the users’ browser will always show the address of the linking webpage or website, but the linked contents remain stored on the server of the linked webpage. Currently, deep linking is implemented mostly by means of:

 

a.       image links;

 

One points to its own webpage an image (or an article, or part of music or video) from another person’s webpage, using the IMG command, so that the linked contents will show on the user's computer screen, as if it is part of the linking webpage that the user is viewing.

 

b.       frames.

 

The framing technology allows a webpage to be divided into a few independent frames, which display different contents from different servers at the same time and roll separately. Using this technology, one can display on his own webpage any contents he requires from others’ websites, and block out any other unnecessary contents, such as, names or advertisements of such other websites. A user who browses a website containing frames will not know that the contents in frames are not owned by the website. The biggest distinction of a deep link from a common link is that users can hardly perceive the linked contents as from another website. This may lead to disputes between the linking party and the linked party for unfair competition, and disputes between copyright holders and the linking party for infringement of network dissemination right.

 

For the dispute between copyright holders and the linking party, a few scholars argue that deep linking to copyrighted works constitutes direct transmission of the copyrighted information and infringement of the network dissemination right of the copyright holder, mainly on the ground that although the linking website does not store any copyrighted contents, it is deemed as the provider of the copyrighted contents by users who, due to deep linking, can hardly perceive the copyrighted information as from another website to which the deep links point. This is the “user perception criteria”.

 

We argue that a deep link, as any other links, does not upload or copy the contents it links to, and deep linking to copyrighted contents is not to “offer” it to the general public for the purpose of network dissemination. Therefore, deep links should be dealt with by the server criteria. Also, if a provider of deep links to copyrighted contents were determined to be a provider of such copyrighted contents on the ground of user perception, the provider would be responsible for stopping the infringing act, that is, to delete the infringing contents. But, as a matter of fact, it is unable to do this, and what it can do is to delink with the copyrighted contents. This leads to a contradiction between infringement determination and civil liability. Also, the server criteria are based objectively on the fact of uploading, while the user perception criteria are based subjectively on the source of infringing contents. Whether an ISP is also an ICP is a factual issue, and the introduction of the subjective user perception criteria into the determination of an objective fact will lead to confusion in court judgment. If a provider of deep links to infringing contents sorts and arranges its deep links in a certain way to guide users, this may be evidence that in subjective terms, the provider “knows or should have known” the possible infringement, but cannot deny that it is the linking service that the provider provides.

 

To sum up, a provider of deep links to infringing contents is not a direct infringer, because it does not act to transmit infringing contents; but it may be an indirect infringer, because its acts contribute to direct infringement. If it is faulty subjectively, it should be responsible for infringement liability; otherwise, it should be responsible for delinking with the infringing contents upon notice by a copyright holder, but not for paying damages. This above rule has been indicated in the relevant cases of the Supreme People’s Court.

 

 

When searching for web pages, a search engine will create on its server a backup or cache of the HTML codes or thumbnails of an indexed webpage on a website, and update such backup or cache at times, according to technical arrangement. If a user clicks on a “cached” link on the page of search results, it will enter the backup of the actual webpage on the server of the search engine. Currently, mainly two types of web caches are involved in copyright disputes. One type is cached web pages, such as those provided by Google or Baidu in their searched results. The contents or titular information of a cached webpage shows when the cache is stored, with a reminder that this is only a cached webpage that the search engine has obtained from an original website. The cached webpage may contain only texts other than images or flashes, and/ or thumbnails of the original images. The other type is cached lyrics, such as those displayed by the Music Box of Baidu when it plays songs. The lyrics are snatched from third-party websites and stored on the server of the Music Box, and the URLs of such third-party websites are often displayed on the player screen. Caching, as a special technical service of a search engine, copies original searched web pages and creates a backup of them on the server of the search engine, as a supplementary information location tool. Caches, as a mirror of the original searched web pages, live and die synchronously with the original contents. We argue, therefore, that although the caching service directly provides third-party contents by copying and storing it on the caching service provider’s own server, it remains an information location tool in general; and that a caching service provider provides an Internet service which helps transmit related information. Generally, the constitutive elements of copyright infringement of a caching service provider are no different from those of contributory infringement of network dissemination right.

 

However, should the exemptions for ISPs as provided for in Articles 22 and 23 of the Ordinance for the Protection of Information Network Dissemination Right be applied to, or should any new exemptions be established for, caching service providers, when determining infringement liability of a caching service provider? We argue that after studying and distinguishing between different caching service operations, two conditions can excuse a caching service provider: a). the cached web pages can never supersede the original; and b). the caching service provider is not faulty subjectively.

 

1.       The cached web pages can never supersede the original.

 

The caching service is technically valuable, as it supplies an effective backup solution, when the original webpage cannot be displayed or can hardly be opened due to slow linking speed, or the server of the original website is temporarily disrupted, or the traffic is clogged, or the link has been redirected. In these cases, not all the information on the original web pages is backed up in the cache. For example, Baidu states in its disclaimer, “Baidu will store a backup in the txt format for each of the web pages it has searched, and this backup is known as Baidu Cache. You can view the contents of a webpage quickly by using its cache. However, as only txt contents is kept, all the other information, such as, any images or music, which is not in the txt format, will be called directly from the original webpage. If the original webpage is not accessible, the non-txt contents on the cached webpage, if any, will not be displayed.” Here, a cached webpage has the following features: (1) any contents of the cached webpage is sourced from and subject to the original webpage, and the caching service snatches original contents without prediction or differentiation; (2) for convenience, a cached webpage keeps only part of the information or thumbnails of the images, other than all the contents on the original webpage; (3) as a cached webpage is made regularly at reasonable times, any change, deletion or blocking occurring to the original webpage will be applied to its cache without unreasonable delay; (4) users are reminded in a reasonably noticeable manner that a cached webpage is a backup created from a third-party webpage, and the URL of such a third-party webpage is provided.

 

We argue that a cached webpage which has these features can never replace the original, for it cannot provide the entire original contents. The caching service is only a supplementary information location tool, and the caching service provider should not be subject to infringement liability.

 

On the other hand, however, the cached lyrics, such as those of the Music Box, have the following features: (1) a cached lyric provides the entire version that fully meets the need of the users of the Music Box, and the users will not have to access the original webpage containing the lyric; (2) the Music Box provides cached lyrics created from certain other than all the large number of original lyric webpages, that is, different from the usual search engine technology which snatches original web contents automatically and indiscriminately, the Music Box screens out certain original lyric webpages; (3) a cached lyric is closely connected and synchronously provided with a song played by the Music Box, which functionally meets users’ need for a music work; (4) when the lyric on an original webpage is changed, deleted or blocked, the Music Box will not synchronously change, delete or block the cached lyric, but go on to search through other cached web pages for the lyric missing, so that all the lyrics are in position under the Music Box. We argue that a cached lyric can replace its original webpage in function, and the cached lyric provider should not be excused.

 

2.       The caching service provider is not faulty subjectively.

 

We argue that the business mode of a caching service provider should be studied in the determination of its subjective fault. For example, the Music Box, which disregards the permission of copyright holders, specifically provides a large number of cached lyrics to attract traffics. It is obvious that such a caching service provider is faulty subjectively.

 

 

In live webcasting, an ICP plays the programs that traditional radio or TV media play. In timed webcasting, an ICP plays such programs at fixed times according to an earlier schedule. Both live and timed webcasting services are very usual ways of transmitting information on the Internet. However, does an unauthorized webcasting infringe the right of a copyright holder or neighboring rights holder? This question constitutes the central and difficult issue in the application of relevant laws. For example, in CCTV International Co., Ltd. v. 21CN Co., Ltd. for copyright infringement, CCTV International exclusively owned the network dissemination right to a program called “Olympic Fire at Mt. Everest”, and 21CN broadcasted the program live and provided replays on its website. In this case, 21CN engaged in both live webcasting and network transmission of the program. We argue that three guidelines should be applicable in the application of laws regarding webcasting service providers:

 

1. The essential difference between webcasting and network transmission is that the former does not allow the general public to obtain a work at a time they choose. Therefore, webcasting cannot constitute infringement of network dissemination right. According to Article 10.1.12 of the Copyright Law, as one of the essential features, network dissemination right allows a user to obtain a work at the time or place that he chooses. That is, different users may obtain the same piece of work from the transmitter’s server at different times. On the contrary, in live or timed webcasting, the time to play a program is arranged by the webcasting service provider, so that the general public can only view, without choice of time, the same part of the program at the same point of time.

 

2. The medium and object of webcasting are essentially different from those of broadcasting. Therefore, webcasting cannot constitute infringement of broadcasting right. The provision on broadcasting right, as in Article 10.1.11 of the Copyright Law, has originated directly from Article 11b Item (1) of the Berne Convention for the Protection of Literary and Artistic Works. Webcasting is different from broadcasting as follows: First, the transmission media are different. The former is transmitted via the Internet, and the latter via radio stations or TV stations. According to the Berne Convention, broadcasting right can be exercised basically by wireless means, and wired means is used only in rebroadcasting broadcasted works. Webcasting, live or timed, can be done by both wired and wireless means. Second, the objects are different. Webcasting transmits digital information, while broadcasting transmits wireless radio waves.

 

3.       Webcasting of works and broadcasting of audio/visual products are governed by different laws. For a copyrighted work, Article 10 of the Copyright Law provides for 17 specific rights under copyright, and the last item is the all inclusive “other rights”. As stated, live or timed webcasting does not constitute infringement upon network dissemination right or broadcasting right, and in the case of webcasting infringement, copyright holders are protected under Article 10.17 of the Copyright Law. For an audio/visual product, however, the producer is not granted broadcasting right by the Copyright Law, and there is no such an all inclusive clause on “other rights”. Therefore, due to the absence of legislation, the producer of an audio/visual product cannot seek for infringement liability against unauthorized providers of the product through webcasting. From the legislative intention and standpoint of the existing Copyright Law, copyright holders enjoy more protection compared to neighboring rights holders. With the time passing and the network technology evolving, the legal protection for neighboring rights holders will have a breakthrough in the respect of non-interactive network transmission of their products.