A proposed broad copyright exception for text and data mining that favors AI developers may not be welcome news for rightsholders.
On June 28, 2022, the UK Government published its response to its consultation on “Artificial Intelligence and IP: Copyright and Patents” (Response) that started in October 2021.
Among other things The government has announced its intention to introduce a new copyright and database exception that will allow text and data mining (TDM) for any purpose, provided that the party using TDM is given lawful access to the material.
TDM is the use of automated computational techniques to analyze large amounts of information to identify patterns, trends, and other useful information. TDM typically requires copying of the material to be analyzed, which requires obtaining a license from the rightsholders or invoking a legal exception.
Rightsholders would not be able to object to this proposed exception, but they would still have safeguards to protect their content, including a requirement for lawful access, allowing rightsholders to charge for access.
TDM plays a significant role in the development and training of AI systems and applications. In line with the goals of its National AI Strategy 2021, the UK Government hopes that the proposed exemption, which will be weighted in favor of TDM users such as businesses and researchers, will help strengthen the UK’s position as a leading global player in the field to secure by AI.
The proposed exemption would make it easier to deploy TDM in the UK, but whether relevant rightsholders would respond by introducing or increasing restrictions on access to their content, or even choosing not to publish content at all, remains to be seen.
UK law already provides a copyright exception for TDM under Section 29A of the Copyright, Designs and Patents Act 1988 (CDPA), which was introduced in 2014. This provision is subject to a number of limitations, including that the purpose of the TDM must be non-commercial research, and that the researchers must have lawful access to the material.
This existing TDM exception relates only to copyright and does not apply to the SUI generic Database right existing under UK law. This right arises when significant investments have been made to maintain, verify or present the data in a database. This right protects against the extraction or re-use of all or a substantial part of the content of the database (which may involve the repeated and systematic extraction or re-use of non-essential parts of the content of a database). Therefore, all database TDMs that qualify for this database right currently require a license from the respective rights holder.
The UK Government’s proposal to introduce copyright law and Database exception allowing TDM any Purpose, with no rightholder opt-out, was the most user-friendly of the options discussed during the consultation. The response indicates that the more balanced options proposal was adopted on the basis that it will support AI and broader innovation and bring benefits to a wide range of UK stakeholders.
The government’s proposal would go beyond the TDM exception introduced by the EU in 2019 as part of the Copyright in the Digital Single Market Directive (allowing rightholders to opt-out of the exception, except in relation to scientific research), and is comparable scope of the broad TDM exception under Japanese law. US law is considered similarly favorable to TDM, although it relies on the “fair use” doctrine rather than a specific TDM exception.
While rightholders could not opt out of the proposed exception, the requirement of lawful access means they can choose the platform on which to make their works available and whether and on what basis to charge for access. Rights holders will also be able to take steps to ensure the integrity and security of their systems.
In particular, the proposed exception should not allow this any Copying copyrighted works or extracting from protected databases for any Purpose. For example, the current exception for non-commercial TDM is limited to making a copy of a work in order to perform computer analysis of everything recorded in the work. Giving this copy to another person or using the copy for any other purpose without the permission of the relevant copyright owner is still infringing conduct.
Analysis and Impact
This proposal is a favorable development for companies, developers and researchers involved in AI. In contrast, it’s probably a less welcome development for rightsholders, as they can’t opt out.
Prior to draft legislation and implementation, respective rightholders should review the access regimes for their core copyrighted works and databases, and consider implementing technical safeguards such as subscription bans and application programming interface (API) restrictions where appropriate.
Depending on the commercial implications of the proposed exemption, some rightsholders may well choose not to publish their content at all.
Notwithstanding that TDM activities may also raise privacy issues, the government’s response stressed that privacy was not the subject of the consultation. Therefore, organizations affected by these developments must separately consider the privacy implications of TDM activities, even if they fall within the scope of the proposed exemption.
 The UK government has also indicated that it does not see any need for legislative changes to remove copyright protection for computer-generated works without human authorship, or to allow patent protection for inventions developed by AI.