An Open Letter to UNESCO on “A Draft Text of a Recommendation on the Ethics of Artificial Intelligence”
by Neil Turkewitz
Let me begin by observing the wisdom of this undertaking, and expressing my appreciation to UNESCO for its vision in examining the development of AI from the standpoint of its impact on culture — viewed both within a narrow framing of the protection of cultural artifacts, and more broadly on the culture and well-being of societies. It is essential that fundamental issues of ethics and culture be considered at an early stage in the development of AI, rather than seeking to implement “solutions” at a later stage when expectations and modalities are already developed and effectively hardwired from an engineering and societal standpoint. I also greatly appreciated and would like to associate myself with the observation of how AI will have both quantitative and qualitative impacts on society, with the possibility of either ameliorating or aggravating existing inequality and bias. I highlight in particular the following:
“The research, design, development, deployment, and use of AI systems should respect and preserve human dignity. The dignity of every human person is a value that constitutes a foundation for all human rights and fundamental freedoms and is essential when developing and adapting AI systems. Human dignity relates to the recognition of the intrinsic worth of each individual human being and thus dignity is not tied to national origin, legal status, socio-economic position, gender and sexual orientation, religion, language, ethnic origin, political ideology or other opinion.
This value should be respected by all actors involved in the research, design, development, deployment, and use of AI systems in the first place; and in the second place, be promoted through new legislation, through governance initiatives, through good exemplars of collaborative AI development and use, or through government-issued national and international technical and methodological guidelines as AI technologies advance.”
I also want to note my strong agreement with the following: “It should always be possible to attribute both ethical and legal responsibility for the research, design, development, deployment, and use of AI systems to a physical person or to an existing legal entity. Human oversight refers thus not only to individual human oversight, but to public oversight. It may be the case that sometimes humans would have to share control with AI systems for reasons of efficacy, but this decision to cede control in limited contexts remains that of humans, as AI systems should be researched, designed, developed, deployed, and used to assist humans in decision-making and acting, but never to replace ultimate human responsibility.”
There is a great deal of anthropomorphism in discussions of machine learning and AI, diverting attention away from issues of accountability and undermining a true understanding of agency. For AI to advance the public interest, it must reflect the values of the societies in which it operates. We must not assign a false agency based on some notion of tech manifest destiny to the development of AI. This is a matter of the first importance, and I fully embrace the framing put forward by UNESCO here.
Turning to the more discrete issue of the impact of AI and cultural production and preservation of diversity, I must say that I was quite disappointed. In a framework for the development of AI from an organization focused on Culture, the lack of any meaningful discussion of how AI may affect the creation of cultural works, the preservation of cultural diversity, or the status of cultural workers is more than mildly distressing. There are so many important intersections of cultural production and AI, yet these go wholly unrecognized or discussed. While, given the role and jurisdiction of WIPO, I do not expect UNESCO to develop copyright norms in relation to AI, the failure to highlight the importance of these issues in cultivating the growth of an ethical framework for the development of AI is mystifying. It is particularly odd given the reference in the document of Article 27 of the Universal Declaration of Human Rights. (1948) emphasizing the right to share in scientific advancement, while wholly ignoring the second paragraph which provides: “Everyone has the right to the protection of the moral and material interests resulting from any scientific, literary or artistic production of which he is the author.” I call on UNESCO to rectify this very material omission in subsequent drafting of this document.
As the document stands, there is really only one observation about AI and copyright, and it is both ambiguous and not central to the relationship between the two. I refer to the following: “AI has implications for cultural identity and diversity. It has the potential to positively impact the cultural and creative industries, but it may also lead to an increased concentration of supply of cultural content, data and income in the hands of only a few actors, with potential negative implications for the diversity of cultural expressions and equality.”
While issues of global competition and economic and cultural hegemony are certainly issues that must be considered in the development and deployment of AI, they are far from the only issues presented by AI in relation to culture. From a cultural standpoint, it is not sufficient to think of AI only as an asset that may be deployed to ways that may increase the digital divide. There are much more fundamental questions that need to be addressed — both in the use of existing works to fuel the development of AI, and in the impact on cultural workers as a result of competition between human-authored works and those created through AI, regardless of whether they are substantially similar to existing works or not.
The failure to consider the rights of cultural workers as they relate to the development of AI threatens to undermine the importance that national governments and other multilateral institutions attach to this critical issue as they undertake examination of the relative risks and opportunities of AI. This work must proceed from a proper understanding of the role of copyright in advancing economic development and cultural diversity. I highlight, as noted in my comments to Special Rapporteur in the Field of Cultural Rights, Farida Shaheed, nearly a decade ago:
“Make no mistake — cultural hegemony flows not from the protection of intellectual property, but from its absence…An effective and functional copyright environment is not a panacea; it does not on its own create global parity in the marketplace of ideas. But it does give individual creators a fighting chance, and an opportunity to compete. The ability to generate revenue from one’s creativity — to earn a living as a creator — is central to a society’s ability to foster cultural production. In its absence, dreams and creative lives perish. The moral and economic aspects of this equation are inseparable. We simply must ensure that all creators, regardless of their location, are able to enjoy the fundamental human right to choose the manner in which their creations are used as reflected in international law…By permitting creative genius to be fueled by market forces, we unleash the cultural power and potential of the diversity of individuals, freeing creative impulses from the tyranny of centralized controls and making creative works accessible to the public at large. While copyright may be inadequate on its own in creating fair market conditions, it remains by far the most powerful tool for fostering creativity and democratizing culture itself.”
I do not intend to here lay out all of the copyright issues that I believe need to be considered in examining the relationship between copyright and AI. Suffice it to say that there are at least two major threads — the use of pre-existing materials to train AI, and the exploitation of works created through AI in competition with more traditionally created works. What I will do here is to set out a broad framework for the consideration of these issues. I begin by observing the importance of resisting any decontextualization of the consideration of these issues, and to ensure that our thinking about AI and copyright reflects broader issues in society related to the effects of automation and the use of algorithms which increasingly define the parameters of our existence, and which, as expressed brilliantly by scholars like Ruha Benjamin (Race after Technology) and Safiya Noble (Algorithms of Oppression) can reify and reinforce existing inequalities in power and produce even greater injustice.
If current narratives are to be believed, the future of writing, singing, composing etc. will increasingly be in the hands of machines. Now of course, machines don’t have hands, but nor do they have creativity. The works ingested by the machine are the raw data by which the machine becomes capable of reconfiguring words, symbols, notes etc. into new works. They are not “reading” as such — a point I highlight here because it has copyright implications as well as moral ones. Some observers liken machine “learning” to how a human might ingest a book, combing through the protected expression while retaining the unprotected ideas. But while a human might very well operate in that manner, it’s a terrible stand-in for the operation of machines which by their very nature “learn” through reproduction, with such reproductions forming the basis of any new output. Those reproductions of expression, even if temporary, are the raw materials used for the development of new forms of expression. In other words, AI isn’t just inspired by the works it ingests — it owes its very existence to them. As such, any theory which suggests that ingested works lack economic or cultural significance couldn’t, in my view, be more incorrect.
AI is the distillation of that which went before, and as such, depends on the past for all of the potential value it may create. Copyright includes the right to create derivative works. AI is, by definition, the very essence of a derivative work. It may be that fair use or other exceptions at national law will permit the unpermissioned, uncompensated use of certain preexisting materials for machine learning in certain instances, but let’s undertake that analysis with open eyes and without resort to theories that would permit massive commercial use as long as the output isn’t substantially similar.
In conclusion, I urge UNESCO to reflect the importance of paying attention to issues of competition that will affect cultural workers. UNESCO should use its platform to warn against any narrow framing that would allow policymakers to consider this as a purely technical matter, devoid of human consequence. Allowing the unauthorized use of images and other content to inform the development of AI is part and parcel of the erosion of human agency and would represent a form of indentured servitude for those contributing to a future beyond their control. And perhaps even more fundamentally from a copyright standpoint, it would force creators to “educate” a system that theoretically will remove the basis of their livelihoods. Most importantly, we need to remember that we are constructing the rules for an emerging digital society. If those rules fail to give effect to core values of personhood and choice, we will be facilitating the cultivation of a dehumanized and dehumanizing society.
