Assignment 2 Final
Aidan Ward
10/24/2023
Heather Christy Robinson
ENG 160
The Importance of Regulatory Methods Regarding Artificial Intelligence
While this could become an incredibly positive force in human progress, the potential negatives bring thoughts of dystopian novels to mind. Due to this paradigm, and the profound consequences of a regulatory inaction, some form of oversight is needed. Although not solving the root of the problem, one of the most compelling methods of regulation currently on the table is that of copyright. Due to the nature of Generative Artificial Intelligence (G.A.I.) requiring copious amounts of unique human-created inputs, the originality of generated content is called into question. This paper will firstly address the status and scope of copyright, denoted by current legislation and precedent. Furthermore,
Copyright law is designed to protect an individual’s intellectual property, whether it’s music, art, literature, or other art forms from theft (Lamb). However, within copyright law lays a caveat titled “Fair Use”, which allows use of copyrighted materials under specific conditions; specific conditions that facilitate a transformed product in some capacity. The intentional subjectivity of “transformative” can be largely denoted to one of two categories: commentary and criticism, or parody (Richard). Both arenas of transformative use involve using the copyrighted material as reference content, and creating substantive changes or use-cases that avoid de-valuing the content. The implications this has for G.A.I. are cloudy, as the law was clearly not written around this new technology, yet it is the most applicable regulatory mechanism in contemporary American law. This fact, paired with the inherent expansive nature of the concept of transformation necessitates legal precedent to curate adequate applications to generated content. While the explicit details of this G.A.I. dilemma lay in uncharted waters, similar issues have been explored in past cases like Authors Guild v. Google and Andy Warhol Foundation for the Visual Arts v. Goldsmith (Jassin). The former outlined an acceptable case for G.A.I. to be used in Google’s processing of books into their searchable databases, allowing individuals to read small portions of books without consent of the publisher or author. While not exactly applicable to the current state of G.A.I., it has led to a “wild west” mentality of intellectual property, incentivizing technology companies to disregard any potential copyright claims. Contrary to this is the aforementioned case of Andy Warhol Foundation for the Visual Arts v. Goldsmith. This recent case established precedent that if an original work is significantly harmed economically, even if it would otherwise be qualified as Fair Use, Fair Use status is not applicable. With the increasing prominence of G.A.I. content, the oversaturation of the market would lead to inherent devaluation. This may prove a useful outlet for future litigation regarding copyright and G.A.I., but does not directly implicate any G.A.I. firms, rendering it ineffective. These precedents illustrate that copyright may be a potential outlet of regulation but are not currently prepared to address these novel issues. Because of this, specific regulation is further required.
These issues of non-regulation hit especially close to home for journalists. Journalism is predominantly dominated in a text-based medium, meaning the proliferation of text-based generative A.I. such as Chat-GPT poses an existential threat to the journalist profession. This view is echoed throughout the entire journalist community, with the effects of early automation efforts being felt by mass layoffs. Early adopters, such as Microsoft (A company whose investment in Open A.I., the creator of Chat GPT caused a rejuvenating stock rally in the fourth quarter of 2023, nearly becoming the most valuable stock in the world(Karaian)), have begun implementing artificial intelligence on their website. This change will influence the millions of Americans who utilize windows products, and as such are prompted to use one of the world’s most influential news sources (O’Sullivan). The labour aspect is not the only reason behind the dismay of the proliferation of G.A.I., as the quality of reporting also suffers. This new journalistic revolution has been plagued by fake news, unreliable sources, and self-reported “low quality articles” (O’Sullivan). These issues call for systemic reform, as current oversight is being outpaced by the innovations in this sector. Logically, competing news organizations must innovate in step, with some attempting to integrate artificial intelligence in a productive yet controlled manor. Such is the case with the Associated Press, a news organization that has long sat on the peak of journalistic integrity and decorum is actively leading a movement for reform. The A.P. has established a set of internal regulations that regulate the use of A.I., restricting any instances of content generation, instead utilizing A.I. as unvetted source material (David). This effectively means journalist treat any information sourced by G.A.I. under a similar protocol to that of a whistleblower, ensuring validity and accuracy. While this is a net positive, it does not go far enough. A wide coalition of news organizations recognize the unsustainability behind this method of regulation and are pleading for transparency reforms, explicitly stating that “these technologies can threaten the sustainability of the media ecosystem” (Preserving public trust in media through unified AI regulation and practices). This begins to illustrate the potential issue with non-regulation, but darker clouds are on the horizon, threatening the solvency of international democracy and cooperation.
As illuded to in the beginning of this essay, the ramifications of a free-reign development philosophy may create new, complex issues. Potential issues include both the individual level threats of labour rights, as discussed, alongside the state level concerns involving constituent based public policy decisions. This becomes obvious when we examine how decisions in government are made. Decisions are based on consensus, and consensus is thus derived from informed viewpoints constructed through independent research, most often through reliable journalism. A core tenant of civil society is the participation in government by its people, and while state actors may have the knowledge and resources to avoid falling into this trap of accidental ignorance, the voter base who elects these actors to power may not have this same knowledge. To avoid this trap and to ensure a trustworthy and integrous news-ecosystem, we must achieve a balance of algorithmic utilization and ethical morality (Ahmad). Within the last year, we have seen a proliferation of low-quality article content farms that tip the scales towards an algorithmic dependence (Brewster). Many of these articles are plagiaristic in nature and lack a manual review process (O’Sullivan). While I contend that this unregulated space is ultimately unhealthy, others argue that we are merely in a state of development that must be explored first and then regulated.
This idea of delayed regulation is parroted by titans of the tech industry, such as the “Father of ChatGPT” Sam Altman (Bello). During Altman’s development of the intelligence framework, he has operated under a mentality reminiscent of the “move fast break things” mantra that Mark Zuckerburg once coined (Nolan). Altman has described his development goals as an “A.I. arms race” against competing startups, a sentiment echoed by its business partner Microsoft in an official memo. This negligent attitude was explicitly stated by Sam Schillace, Microsoft’s chief technology officer with the statement it would be an “absolutely fatal error in this moment to worry about things that can be fixed later.” (Weise) The rationale behind this decision is to compete with other firms attempting to harness similar technology, a methodology akin to Facebook during its inception. Although likely a profitable move, this has historically been a harmful approach as was the case with Facebook. Facebook notably suffered numerous crisis due to their “move fast and break things” mantra, with the Cambridge Analytica scandal being the most notable (Nolan). While we cannot rewrite the mistakes of our past, I believe it is our responsibility to prevent history from reoccurring, especially considering the elevated stakes that come with this new technology.
These issues as discussed are sure to change, as the dynamic landscape we find ourselves in is subject to the ebbs and flows of technological advancement alongside potential government oversight. While there is a distinct lack of government intervention at the current moment, we have past precedents regarding copyright that may prove to be effective in regulation of this novel technology. While current Fair Use statutes and the law at large are unprepared for this new age of technology, the potential avenues of regulation are written into the law and may be levied to provide guard rails. These guard rails have proven necessary, as the current climate projects a grim outlook on the future of journalistic integrity and morality for many major news organizations. This technology has further proven to be a potential threat to exercises of civil society, and we should not trust those who are beholden to capital to independently regulate their own potential gains. As we have seen with countless instances in the past, if left to their own behest, capital corrupts, favouring expansion over empathy. For this reason, I hold that further regulation is of prescient concern to all individuals who interact within society at large.
Works Cited
Stim, Richard, and Rich Stim Attorney at law. “What Is Fair Use?” Stanford Copyright and Fair Use Center, 25 Nov. 2021, fairuse.stanford.edu/overview/fair-use/what-is-fair-use/.
What Is Fair Use? Is an article that provides a basic outline around the qualifications and restrictions of Copyright Fair Use Law. The article defines the primary qualification of Fair Use Law, being a transformative altering of the content. The article further discusses the use cases of fair use law, as a defense against copyright infringement. The article offers an objective, easily digestible definition of Fair Use. The article’s most useful application for my case is to establish that fair use is only applicable to transformative content that implements commentary and criticism or parody. This outlines an adequate, objective interpretation of fair use law, an interpretation that will lend itself as the bedrock for the rest of my paper. By utilizing these ideas, we can evaluate the current status of copyright law for further analysis.
David, Emilia. “The Associated Press Sets AI Guidelines for Journalists.” The Verge, The Verge, 16 Aug. 2023, www.theverge.com/2023/8/16/23834586/associated-press-ai-guidelines-journalists-openai.
This article outlines the evolving landscape of journalism in respect to the rapid proliferation and advancement of artificial intelligence. The article explains how in a landscape dominated by written articles, a generative artificial intelligence poses prescient concerns about the future of the profession of journalism. Furthermore, alongside the mentioned lack of regulation, it introduces the Associated Press’s newly created a set of internal regulations on the use of artificial intelligence in journalism. The article is fairly pro-journalist, as it introduces the newly mentioned regulations, but it also doesn’t completely disavow use of artificial intelligence either, landing in a slightly more centrist camp. This proves useful in the greater landscape of my paper, as it can outline a more realistic approach to regulation. Although ultimately insufficient in the long run, it can provide insight on the ideology and outlook of the journalist’s perspective.
Lamb, Anna. “Martin Puchner Traces History of Cultural Appropriation.” Harvard Gazette, 9 Nov. 2023, news.harvard.edu/gazette/story/2023/03/martin-puchner-traces-history-of-cultural-appropriation/.
The Harvard Gazettes’ “Culture belongs to everyone (and no one)” outlines the writings of Martin Puchner and his research on the genesis and evolution of culture throughout the ages. In the article, Puchner discusses an alternative notion of culture. A notion of culture that is institutionalized rather than hereditary. He further discusses the cyclical nature of culture, and how it manifests itself with re-occurring trends and the unoriginality of cultural phenomenon, outlining his belief of the inability to formulate original content. Puchner concludes by cautioning against an ouroboros of disdain regarding past cultural digressions, arguing that the same scrutinization will be placed on us in the future for our ignorant ambivalence. Puchner seems to take a much more academic approach to the topic, not particularly landing on either side of the matter. This becomes useful when discussing the ambiguity of culture and originality, an idea utilized in the paper.
Jassin, Lloyd J. “Generative AI vs. Copyright: An attorney examines what artificial intelligence will mean for the publishing industry.” Publishers Weekly, vol. 270, no. 39, 25 Sept. 2023, p. 96. Gale General OneFile, link.gale.com/apps/doc/A768425332/ITOF?u=nysl_se_sojotru&sid=bookmark-ITOF&xid=6c8d2b33. Accessed 3 Nov. 2023.
This piece titled An attorney examines what artificial intelligence will mean for the publishing industry attempts to convey the current landscape of generative artificial intelligence in regards to copyright law. Specifically discussing the murky nature of fair use laws and how they would apply to GAI, as well as discussing the potential benefits of future regulations. This article references two precedents that have the possibility to apply to generative artificial intelligence (abbreviated to GAI) and develops their similarities and differences to our current situation. This short article doesn’t pick a certain side of the argument per say, instead offering a more objective stance on the possible future of this area of law. This proves relevant in my piece, as it offers current possibly applicable precedent and explains it, as well as the potential holes in the argument that could nullify any possible application.
Gold, Jon. “Generative AI and US copyright law are on a collision course.” ComputerWorld.com, 22 Sept. 2023, p. NA. Gale General OneFile, link.gale.com/apps/doc/A767118872/ITOF?u=nysl_se_sojotru&sid=bookmark-ITOF&xid=2153e2bb. Accessed 3 Nov. 2023.
This article focuses more on the copyrighted nature of the output of generated artificial intelligence, explaining how the copyrighting process cannot be applied unto generated works. The article also discusses other mediums of regulation, illustrating a rather pessimistic view on the future of artificial intelligence copyright law. The article also discusses a lack of multilateral regulation for any medium, stating more research is necessary to create concrete resolutions. This article is clearly antagonistic towards generative artificial intelligence, providing insight on current lawsuits and generally operating with a negative tone. The article declares that the regulation is imminent in some form and will be necessary to foster a healthy ecosystem as has been the case historically. This is the perfect article for my paper, as although slightly short, it describes many of the issues my paper revolves around, specifically advocating for regulation while utilizing historical precedent and copyright law.
Ahmad, Norita, et al. “The News Ecosystem in the Age of AI: Evidence from the UAE.” Taylor & Francis, vol. 67, no. 3, Feb. 2023, pp. 323–52, https://doi.org/10.1080/08838151.2023.2173197.
Bello, Camille. “Who Is Sam Altman? From College Dropout to Creator of OpenAI’s ChatGPT.” Euronews, 21 Nov. 2023, www.euronews.com/next/2023/11/21/who-is-sam-altman-college-dropout-to-creator-of-ai-chatbot-chatgpt.
Brewster, Jack, et al. “How Copycat Sites Use AI to Plagiarize News Articles.” Newsweek, 17 Oct. 2023, www.newsweek.com/how-copycat-sites-use-ai-plagiarize-news-articles-1835212. Accessed 12 Dec. 2023.
Gordon, Donie O’Sullivan, Allison. “How Microsoft’s AI Is Making a Mess of the News .” CNN, 2 Nov. 2023, www.cnn.com/2023/11/02/tech/microsoft-ai-news/index.html.
Karaian, Jason. “Microsoft’s Stock Hits Record High after Hiring OpenAI Outcasts.” The New York Times, 20 Nov. 2023, www.nytimes.com/2023/11/20/business/microsoft-stock-openai.html.
Kurzweil, Ray. The Singularity Is Nearer. Random House Large Print, 2022.
Miller, Scott Gunnison. “Ray Kurzweil’s Prediction Scorecard Now That ChatGPT Is Here.” Medium, 3 Mar. 2023, medium.com/@sgunnisonmiller/ray-kurzweils-prediction-scorecard-7d40ee2ff42a#:~:text=In%20his%202005%20book%2C%20The. Accessed 12 Dec. 2023.
Nolan, Beatrice. “Silicon Valley Has a New Version of Its Beloved ‘Move Fast and Break Things’ Mantra.” Business Insider, 6 Dec. 2023, www.businessinsider.com/silicon-valley-move-fast-and-break-things-sam-altman-openai-2023-12. Accessed 12 Dec. 2023.
Weise, Karen, et al. “Inside the A.I. Arms Race That Changed Silicon Valley Forever.” The New York Times, 5 Dec. 2023, www.nytimes.com/2023/12/05/technology/ai-chatgpt-google-meta.html.