Why We’re Advocating for a Cautious Approach to Copyright and Artificial Intelligence
CopyrightOn 14 February 2020, Creative Commons (CC) submitted its comments on the World Intellectual Property Organization (WIPO)’s Issues Paper* as part of WIPO’s consultation process on artificial intelligence (AI) and intellectual property (IP) policy.
In this post, we briefly present our main arguments for a cautious approach to regulating AI through copyright or any new IP rights.
- There is a lack of clarity and understanding around “artificial intelligence”
Technological developments in AI are fast-paced and raise complex policy, legal, and ethical issues that deserve global attention. However, AI needs to be properly understood before any copyright implications can be addressed. At this nascent stage of AI technology, there lacks consensus on how to define AI.
As discussed more fully in our submission, AI algorithms differ in the depth and breadth of input required to produce coherent output, and it’s not clear how to judge the originality of a work essentially composed of random snippets of thousands or millions of input works. There is also uncertainty about whether and to what extent AI is capable of producing content “autonomously” without any direct human involvement, and whether AI outputs should be protected by copyright. Clarity on these and other basic definitions in the “AI” space is a prerequisite to competent regulation in this arena.
At present, it is ill-advised to force the application of the copyright system—an antiquated system that has yet to adapt to the digital environment—onto AI.
- Copyright is centered on human creativity
Human creativity is the bedrock of copyright. We ought to maintain the expectation that direct human, authorial involvement exists as a pre-condition to determining whether a work is worthy of copyright protection. Outputs from a mechanical process with no direct human involvement should not constitute works protected by copyright absent further examination and understanding of how the many different types of AI operate, generate output, and how closely humans are generally involved in the process. Simply put, it is premature to conclude that all AI applications should be considered “creators.”
AI-generated outputs should not as a default be considered original works. Courts the world over have affirmed that originality is a reflection of the intellectual, creative choices made by the author. The originality bar may be low, but it does exist and must be respected.
Regarding the use of copyright material as inputs in AI applications, Creative Commons’ FAQs clarify how the CC licenses work in the context of openly licensed content that is used to train AI tools.
- New rights would be inappropriate
Copyright is not designed to handle any and all policy issues adjacent to the creation and use of IP. To wit, press publisher rights and the ongoing debate over broadcasting rights demonstrate the danger of overstretching copyright to regulate peripheral issues. Using copyright to govern AI is unwise and contradictory to copyright’s primordial function of offering an enabling environment for human creativity to flourish. Issues such as ethics, privacy, and personality rights, among many others, are valid concerns, but they should be addressed and debated in their respective policy arena, not within the framework of copyright.
Using copyright to govern AI is unwise and contradictory to copyright’s primordial function of offering an enabling environment for human creativity to flourish.
For the same reasons, we strongly urge against the temptation to create new sui generis (specific, tailor-made) rights established for AI-generated content. Incentives and rewards in recognition of the investment made and the innovation brought about by the organizations and individuals involved in the development of AI can be found in other areas, including patents, trade secret laws, and laws protecting against unfair competition.
We suggest proper safeguards if copyright or new rights apply to AI
Assuming WIPO members are keen to rely on copyright or new sui generis rights to regulate AI and protect AI-generated content despite our concerns, this should be done conservatively and with restraint. Members should set a high bar for the creation of such new rights, consider a much lesser term of protection than that provided to the original works created by human creators, and ensure that robust exceptions and limitations are in place to uphold users’ rights, safeguard the public interest, and ensure a vibrant public domain.
Next steps
Creative Commons looks forward to the next version of the WIPO Issues Paper and will continue to take part in WIPO’s discussion process on AI and IP.
To learn more about CC’s work with the World Intellectual Property Organization (WIPO), check out this Wiki page.
*The WIPO consultation aims to gather submissions to identify important issues related to AI and IP, which will form the basis of structured discussions in the future. For example, the consultation seeks answers to questions such as:
- Should AI-generated outputs be protected by copyright?
- Should there be specific exceptions to allow the use of copyrighted material by AI applications?
Posted 20 February 2020