In an era of rapidly evolving (AI) and (ML) technologies autonomous design, music composition, art creation and predictive analytics, bear witness to the pervasive influence of AI and ML in virtually every facet of our daily lives. The seamless integration of (AI) and (ML) technologies across diverse industries has sparked a paradigm shift in the way businesses operate. While these advancements present opportunities for innovation, cost-effectiveness, and efficiency, they simultaneously pose complex legal challenges in the realm of intellectual property laws.

Redefining Copyright and Authorship

The traditional bastions of patents, copyrights, and trade marks are now being confronted with novel scenarios, courtesy of AI and ML algorithms that have become instrumental in creating, analysing, and birthing novel works, processes, and innovations. As the line between human-generated and AI-generated content begins to blur, the question of ownership and authorship looms large, demanding a fresh and astute examination of our existing IP frameworks.

AI, Copyright, and Complex Legal Questions

The Copyright Act, 1978 of South Africa governs copyright and allows for the right to control the use and distribution of certain works. One does not have to register copyright in South Africa, as it automatically exists the moment an original creative work is in a material form. Copyright generally vests in the author of a work subject to certain exemptions. 

In a prior United States of America (US) copyright case where a monkey named Naruto, took a selfie on a photographer’s camera, the question arose of who owned the copyright in the selfie. The US District Court held that the concept of authorship under the US Copyright Act cannot be defined to include non-human animals. Much the same, AI generated works, would possibly (according to the SA Copyright Act) not be considered as works created by human authorship and therefore be unlikely to be protected under copyright. The making of creative works by AI therefore challenges conventional copyright principles and addressing the role of human involvement in such works becomes crucial to maintaining copyright integrity. 

The intersection of AI and copyright law manifests a number of different problems. First, is the issue of how and from what/where AI tools learn, which raises the question as to whether the learning process is infringing copyright.  AI companies scrape images/music/data etc from the Internet and use them to program their AI with different themes, moods and styles. Whether the outputs that AI tools produce are infringing works, would involve an assessment of whether any part of an original work was copied, and if that forms a ‘substantial’ part of the original work. Those who utilise AI to create works therefore need to consider whether they are using any party’s work which may be copyrighted.

For example, in the US, Getty Images instituted an action against Stability AI, alleging that their copyrighted images were copied in the process of training Stability AI’s image generator tool “Stable Diffusion” without their consent, and that this copying would infringe any copyright subsisting in the images.

On the other hand, in trade mark protection, AI’s involvement in branding necessitates evaluating the distinctiveness of AI-generated marks and potential consumer confusion. AI’s reliance on extensive datasets also raises data privacy concerns, requiring a balance between data-driven innovation and individual privacy rights.

Revolutionising IP Enforcement with AI and ML

AI and ML technologies have not merely left their mark on the creation and protection of IP rights but have also instigated a seismic shift in the enforcement of these rights. With the proliferation of digital platforms and the ever-expanding online space, the challenges in enforcing intellectual property laws have reached an exponential magnitude. However, AI and ML offer innovative solutions that bolster the effectiveness and efficiency of IP enforcement endeavours. The scale and intricacy of IP infringements, encompassing counterfeiting, piracy, and online violations, have escalated in the digital age. Conventional manual detection and enforcement mechanisms now find themselves struggling to keep pace with the deluge of online content. In this crucible of challenges lies the untapped potential of AI and ML to revolutionize intellectual property law enforcement.

AI-powered algorithms possess the dexterity to autonomously scour the vast expanse of the Internet, unearthing instances of IP infringements with precision. These algorithms can deftly identify counterfeit products, unauthorized distribution of copyrighted content, and trade mark violations. Their swift analysis of voluminous datasets, coupled with pattern tracking, facilitates proactive and targeted enforcement actions. This technological marvel aids IP owners, enforcement agencies, and digital platforms in swiftly detecting and addressing IP infringements, thereby mitigating the deleterious economic consequences of illicit activities.

The Challenge of Public Domain and AI Works

On a general note, individuals are also at risk of unknowingly agreeing to having their images freely used by third parties as they partake in online challenges such as on TikTok by uploading pictures of themselves to produce an AI generated version. These works, according to US Copyright Laws, are ineligible for copyright protection and accordingly fall within the public domain to be used by anyone for commercial or non-commercial purposes. Copyright protection is therefore not afforded for unauthorised use of such images by third parties. 

More recently, Chat GPT has also been called into question as lawyers in a South African case utilised Chat GPT to provide cases as references in their arguments. Unfortunately, the information provided by Chat GPT was false and made up. Bringing into question the accuracy of such AI derived works and the importance of fact checking.

Global Cooperation in Addressing AI and IP

Global cooperation and harmonisation of IP laws are vital to address cross-border implications of AI and ML on IP rights, fostering consistency in recognition and enforcement. 

To address the challenges posed by AI-generated works, we may consider adopting specific guidelines for such works, possibly amending the Copyright Act to recognize AI systems as co-authors, provided there is significant human input in the creative process. Another option is to assign authorship and IP rights to the AI programmer or developer, recognizing their role in creating the technology that enables the creation of the work’s. Additionally, SA could consider creating a new category of IP rights specifically for AI-generated works, which would distinguish them from traditional artistic works.  

A well-structured regulatory framework is required to infuse AI and ML technologies with a sense of responsibility and ethics within the domain of IP protection. Policymakers must synergize with technology and legal experts to craft laws that proficiently address the challenges posed by AI-generated content and inventions. The regulatory framework must define inventorship and ownership in AI-generated works, establish a balanced approach to fair use in AI-generated content, and create clear guidelines for assessing inventive steps in AI-generated inventions.

Crafting an Effective IP Strategy for AI

Businesses that aspire to embrace AI and ML technologies must engineer IP strategies to shield their innovations and creations. An effective IP strategy must encompass the identification, protection, and enforcement of IP rights pertaining to AI-generated works and inventions. To unlock this realm of possibilities, businesses should:

  • Conduct IP Audits: Meticulously assess existing and potential IP assets to discern AI-driven innovations and creative works that demand protection. 
  • File for IP Protection: Spearhead the charge by filing patent applications for AI-generated inventions, registering trade marks for AI-driven branding, and adhering to copyright requirements for AI-generated content.
  • Review Licensing Agreements: Scrutinize licensing agreements for AI technologies, ensuring their thorough integration with unequivocal provisions on IP ownership, usage rights, and confidentiality. 
  • Monitor IP Infringements: Embrace AI-driven tools for monitoring and detecting potential IP infringements, which shall confer the power to initiate prompt enforcement actions. 
  • Educate the Workforce: Empower employees and stakeholders through training on intellectual property laws, consistently emphasizing the primacy of protecting AI-generated innovations and content.

Promoting Collaboration and Responsible AI Use

Collaborative efforts and licensing agreements possess the innate power to stoke the embers of innovation while ensuring robust IP protection. Licensing agreements shall serve as the conduit for facilitating the responsible use of AI-generated content and inventions, endowing businesses with access to cutting-edge AI technologies while dutifully respecting the rights of IP creators. Meticulously structured licensing agreements engender win-win scenarios, empowering businesses to chart new pathways of innovation while affording AI developers the recognition and remuneration they so rightfully deserve. Additionally, collaborative research and development projects hold the potential to catapult AI-driven innovation to new heights, culminating in ground-breaking advancements that augur well for society. Such collaborations also imbue a culture of responsible AI use, continually emphasizing the sanctity of IP protection and strict adherence to pertinent laws and regulations.

The Future of AI-Generated Works in South African Intellectual Property Laws

The advent of AI-generated works presents both opportunities and challenges for the South African intellectual property laws framework. By proactively engaging in legislative reform, fostering collaboration and dialogue, and promoting awareness and ethical guidelines, South Africa can effectively navigate these challenges and harness the potential of AI-generated works for the benefit of its community and society at large.

关于 Spoor & Fisher：

Spoor & Fisher is Africa’s largest specialised intellectual property law firm, with African roots and global reach. The firm specialises in all aspects of IP law, including trade marks, copyright, patents, registered designs, anti-counterfeiting, commercial/transactional work involving IP, and litigation in these fields. Spoor & Fisher is ranked in the top band in the latest editions of leading legal directories, both local and international, and has a reputation for pioneering thought leadership and contributions to IP law and academia. Clients have trusted the firm to protect, manage and enforce their IP across Africa and the Caribbean for over 100 years.