Artificial Intelligence (AI)
Dating back to the 1950s when the term was first coined and used by emeritus Stanford Professor John McCarthy, Artificial Intelligence (AI) is defined as “the science and engineering of making intelligent machines.”[1] That is, AI enables users to accomplish cognitive functions and tasks associated with intelligent beings such as perceiving, reasoning, and acting under the control of computational devices.[2] Under the umbrella of AI, there are different subsets such as: machine learning, deep learning, robotics, neural networks, and NLP, which may be used in combination in various applications.[3]
Machine Learning
In particular, machine learning, a subset of AI that enables computer systems to capitalize on data and algorithms to analyze data and make predictions, is reshaping the esports sector.[4] The application of machine learning technology in esports has enhanced interactions and created personalized gaming experiences for esports fans on the one hand, while allowing game developers to leverage the technology to generate more creative contents and drive growth on the other hand.[5] At the forefront of the intersection between AI and Esports are leading European esports organizations like G2 Esports that announced its strategic partnership with German tech start-up, Shikenso Analytics to capitalize on the data-driven AI tools to monitor and track the performance of G2 Esports in streaming and social media channels, thus expanding its brand footprint by providing G2 Esports with an enhanced understanding of its activities in the esports and gaming communities.[6] To this end, Shikenso uses machine learning to evaluate performance metrics which, in turn, could aid the strategic decision-making processes on the part of its clients.[7] Given that sponsorship deals and the integration of streamers are inextricably tied to brand association and building a more robust, engaged community, esports organizations will likely be able to capitalize on AI analytics in creating more targeted sponsorship and partnership offers tailored to the audience on the basis of data analyzing viewership, fan preferences and interests.
Likewise, AI analytics can be used to dissect player statistics, game conditions and strategies by tracking player’s movements real-time and rendering AI-enabled prediction of in-game performances, thereby facilitating esports teams and organizations in training their players.[8]
Legal Challenges
However, the use of AI raises legal challenges stemming from its characteristics. First, because data used to train AI are harvested from players’ gaming profiles (or, in the case of a sponsor or partner, their partnership performance metrics), ensuring compliance with privacy laws will be quintessential given the private and sensitive nature of such information.[9] Esports industry stakeholders must ensure that data subjects (e.g., players, streamers, fans) are adequately informed about the collection of data and that consent has been obtained from them pursuant to relevant legal frameworks such as the GDPR.[10] The newly proposed EU Artificial Intelligence Act will likely add another layer of regulatory complexity as the Act places a strong emphasis on transparency, among other values.[11] This means that to facilitate compliance with EU data protection law, the Act requires providers of AI systems to be transparent about the intended purpose of data collection for personal data to ensure the data subjects are reasonably well-informed.[12] It may be challenging for AI service providers to comply with such regulations via traditional models used to obtain user consent (e.g., cookie banners) given that AI systems analyze exponentially large quantities of real-time data, rendering it unfeasible to obtain consent each time data is collected and processed.[13] Likewise, ethical issues can be raised on the part of users in relation to consent. That is, when providing consent to certain uses of their personal information, users cannot predict how data can be repurposed for another use beyond their envisioned scope of consent, thus rendering their original consent inapplicable.[14]
Second, neural networks are another form of AI acting as a game-changer in esports. Neural networks, inspired by the human brain as it name suggests, creates an adaptive system with interconnected nodes (neurons) resembling the human brain to assist computers in making intelligent decisions.[15] Artificial neural networks consist of three-tiered, interconnected artificial neurons (i.e., software modules) that enable the machine to comprehend unstructured data and solve problems.[16] Specifically, outside data enters through the input layer where it is processed, analyzed, or categorized, subsequent to which the hidden layer analyzes the output and processes it further before passing the processed data to the next layer.[17] Finally, the output layer renders the final result of the processed data.[18] Neural networks play a crucial role in generative AI models where new output is created using existing data.[19] Game developers may capitalize on neural networks to generate varied, unpredictable game environment designs using an algorithm called procedural content generation machine learning (PCGML) trained on existing content.[20] Neural networks in PGML algorithms are trained to augment and generate visual elements in games by auto completing certain parts of games, for instance, by painting images with missing parts.[21]
In-game generative content creation can give rise to issues relating to ownership and authorship of the content generated by neural networks. That is, whether developers that generate gaming content using AI can avail themselves of copyright protection is a crucial issue when using AI. Although case law is still developing in this regard, major jurisdictions such as the United States and the EU are currently of the view that the question of ownership is inextricably tied to originality. In the first-ever groundbreaking judgment relating to AI-generated content, the United States District Court of the District of Columbia found that pursuant to the Copyright Act of 1976, copyright did not attach to visual artwork created by AI based on the rationale that copyright law is limited to “original intellectual conceptions of the author.”[22] The court’s ruling that “human authorship is a bedrock requirement for copyright” aligned with the U.S. Copyright Office’s official view that human authorship is the most fundamental requirement to copyrightability as established in both the United States Constitution and the Copyright Act; accordingly, machine-created artwork cannot be copyrighted.[23] Similarly, in the EU, human authorship is a central requirement for the creator to receive copyright protection. The European Court of Justice has established that recognizable “work” is subject to the following criteria: (i) represent an original intellectual creation of its author; and (ii) only elements expressing such a creation should be considered a “work.”[24] As such, a work is original only if its author’s creative decisions are reflected therein. Hence, the US and EU both take the positions that human authorship is a bedrock principle of copyrightability although with the increasing role of AI in content creation, we have yet to see whether the current judicial stance on centrality of human beings in machine-generated work will shift.
As examined, debates surrounding the copyrightability of AI-generated content underscore the need for clarity on issues of human authorship and originality. Thus, while AI offers immense potential for esports, navigating legal and ethical complexities remains critical for stakeholders in the industry.
[1] Christopher Manning, ‘Artificial Intelligence Definitions’ (Stanford University Human-Centered Artificial Intelligence, September 2020) <https://hai.stanford.edu/sites/default/files/2020-09/AI-Definitions-HAI.pdf> accessed 28 February 2024.
[2] Panel on Computer Science and Artificial Intelligence Naval Studies Board & Commission on Physical Sciences, Mathematics, and Applications National Research Council, ‘Computer Science and Artificial Intelligence’ (1997) 1.
[3] Ekin Keserer, ‘The six main subsets of AI: (Machine learning, NLP, and more)’ (Akkio, 24 November 2023) <https://www.akkio.com/post/the-five-main-subsets-of-ai-machine-learning-nlp-and-more> accessed 28 February 2024.
[4] Nick Heath, ‘What is machine learning? Everything you need to know’ (ZDNET, 16 December 2020) <https://www.zdnet.com/article/what-is-machine-learning-everything-you-need-to-know/> accessed 28 February 2024.
[5] Vinayak S., ‘AI-Powered Audience Growth: How Esports Uses Artificial Intelligence To Connect With Fans’ (Forbes, 26 September 2023) <https://www.forbes.com/sites/forbestechcouncil/2023/09/26/ai-powered-audience-growth-how-esports-uses-artificial-intelligence-to-connect-with-fans/?sh=59e5d86b545f> accessed 28 February 2024.
[6] Karsten Schonauer, ‘G2 Esports Teams up with Shikenso Analytics’ (Shikenso Analytics,3 August 2023) <https://shikenso.com/blog/g2-esports-teams-up-with-shikenso-analytics-for-partnership-measurement> accessed 1 March 2024.
[7] Shikenso Analytics, ‘Sponsorship’ (Shikenso Analytics) <https://shikenso.com/#social> accessed 1 March 2024.
[8] Ivan Simic, ‘How can AI improve esports inside and outside the game?’ (Esports Insider, 22 November 2023) <https://esportsinsider.com/2023/11/ai-esports> accessed 1 March 2024.
[9] Usercentrics, ‘Artificial Intelligence (AI), personal data and consent’ (Usercentrics, 18 August 2023) <https://usercentrics.com/knowledge-hub/artificial-intelligence-ai-act-and-consent/> accessed 3 March 2024.
[10] ibid.
[11] Proposal for a Regulation of the European Parliament and of the Council laying down harmonized rules on artificial intelligence (Artificial Intelligence Act) and amending certain Union legislative acts [2024], 2021/0106 (COD).
[12] ibid [Title IV. Art. 52].
[13] Usercentrics (n 9).
[14] Adam J. Andreotta et al., ‘AI, big data, and the future of consent’ (Springer, 2021), <https://link.springer.com/article/10.1007/s00146-021-01262-5> accessed 10 March 2024.
[15] Amazon Web Services, ‘What is a Neural Network?’ (AWS) <https://aws.amazon.com/what-is/neural-network> accessed 10 March 2024.
[16] ibid.
[17] ibid.
[18] ibid.
[19] Nvidia, ‘Generative AI’ (Nvidia)<www.nvidia.com/en-us/glossary/generative-ai> accessed 10 March 2024.
[20] Adam Summerville et al., ‘Procedural Content Generation via Machine Learning (PCGML)’ (new Jersey Institute of Technology, September 2018) < https://researchwith.njit.edu/en/publications/procedural-content-generation-via-machine-learning-pcgml> accessed 16 March 2024.
[21] ibid.
[22] Thaler v. Perlmutter (United States District Court of the District of Columbia 2023) (No. 22-1564 (BAH)).
[23] United States Copyright Office, ‘Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence’ (U.S. Copyright Office, 16 March 2023), <copyright.gov/ai/ai_policy_guidance.pdf> accessed 16 March 2024.
[24] Case C-683/17 Cofemel – Sociedade de Vestuario SA v G-Star Raw CV [2019] (ECLI:EU:C:2019:721), para 29.
