Home Artificial Intelligence The Fair Use Tango: A Dangerous Dance with [Re]Generative AI Models

The Fair Use Tango: A Dangerous Dance with [Re]Generative AI Models

The Fair Use Tango: A Dangerous Dance with [Re]Generative AI Models

Photo ©️2023 Neil Turkewitz

By Neil Turkewitz

Every yr around this time — designated by some organizations as “Fair Use Week,” the copyright world engages in a rigorously orchestrated dance. Organizations & individuals which can be skeptical in regards to the role of copyright and fearful about its implications for expressive freedoms rejoice the role of fair use in limiting proprietary interests while recognizing the worth of what they see as an appropriately limited set of rights under copyright. For his or her part, copyright owners & folks that see the incentives of copyright as being fundamental to the interests of the general public by creating incentives for the creation & distribution of creative works use the occasion to focus on the risks of expanding fair use beyond its rigorously circumscribed purposes, while recognizing the importance of what they see as an appropriately constrained fair use doctrine as an important feature of the legal & cultural landscape.

Through the years, I actually have been a willing participant on this dance, although I actually have increasingly wondered whether there’s any value in repeating the well-known choreography. Is anyone learning from this annual exercise? Is anyone chatting with folks that don’t already agree with them? I actually have tried on several occasions to achieve this, but I fear that my role as a known champion of creatives & copyright dooms such an effort to search out common ground. Five years ago, I wrote:

”Lost in talk of “copyright wars” and an over-simplified copyright versus tech narrative, is the undeniable fact that there’s an ideal deal of common interest and agreement. Indeed, the age of data could be arid (and uninformative) within the absence of either information or modes for distributing and accessing it. Tech and content need one another, are a part of one another, and actually bleed into each other. While there are actually some outliers and differences related to theories of utilitarianism versus natural rights, at core there’s broad agreement on the foundational issues and purposes of copyright law, including the necessity for a copyright regime that ensures support for working creators and promotes innovation. Most tech firms would agree that copyright protection serves the general public interest, and that copyright systems should foster latest types of expression. The creative sector for its part supports the concept that the establishment of recent, robust and effective protection should include appropriate limitations or exceptions to rights, consistent with international treaties and the needs of various cultures (each legal and artistic).”

I wish I could say that we have now progressed a more common vision within the five years since that appeared, but that might be incorrect, and I’ll leave hallucinations to ChatGPT and other large language models. As an alternative, I’ll offer what I see as a fundamental truth — that no matter where one is on the copyright-fair use scale, it’s now time for us all to be joined in ensuring that the training of publicly available (i.e. non-research) AI generative models (StableDiffusion, Midjourney, DALL-E, ChatGPT, etc) takes place only on the premise of consent, not through nonconsensual text & data mining. That is way larger than copyright, but copyright will, for higher or worse, play a very important role in the development of rules for a digital world (which is to say, the fashionable world) and the extent to which it’s anchored in human dignity and self-determination. My humble suggestion is that we attempt for the “higher” option.

The problems presented by generative AI are unlike any that society has faced before. It forces us to think about what makes us human, and the extent to which we’ll allow the constitutive elements of our personhood to be captured by machine. From our personal photos, to our stories, to our medical data, to…well, every little thing that is understood about us and captured in some manner on the web, is that this all grist for a mill of incomparable appetite? Or do we would like to take steps to be certain that our emerging and ever-changing digital world is built on the premise of clear, freely granted, consent? Not as a part of the non-negotiated & non-negotiable terms of living in a digital world, but as an expression of free will during which the relevant actor has agency?

In lots of respects, copyright offers the cleanest, most immediate way of answering at the least a part of this query. It exists upon creation of an original expression, not depending on any formalities, and making a piece available to the general public doesn’t divest the property interest — indeed, copyright is essentially, albeit not completely, rooted in a desire to incentivize the sharing of 1’s works with the general public. Unlike more complicated privacy issues, there isn’t any have to define the protected interest — it’s already well-defined and guarded globally through quite a lot of international instruments. There is basically but a single issue standing between self-determination and machine supremacy — ensuring that training AI without the consent of the creator will not be excused by exceptions to copyright, including after all, fair use.

Coming back to the premise of this piece, there are numerous complicated cases during which competing equities place the doctrine of fair use front & center, and during which parties may reasonably disagree (or agree) about where & how you can draw appropriate boundaries in pursuit of achieving the Constitutional purpose of copyright. To my mind, the unauthorized use of copyright works to coach AI will not be such a case. And by this, I don’t mean that there are not any legal arguments which may be made to justify such use. I believe there are, although I believe they’re weak. But I do think that there are not any moral arguments to be made to justify this unprecedented appropriation. That latest transformative works could also be produced from a database developed on the back of conscripted labor can have some resonance in law. Nevertheless, it fundamentally fails the humanity test.

Fair use has traditionally been seen as advancing redistributive justice — stopping powerful actors from restricting competition within the marketplace of the expression of ideas. As I wrote last yr during Fair Use Week:

“Fair use is usually portrayed as a lever against monopolies, exercised by those without market power against those that do. But as noted by Ben Sobel in his 2017 article “Artificial Intelligence’s Fair Use Crisis,” “Today’s digital economy upends this narrative. … This pivot in market dynamics should prompt a corresponding shift in attitudes towards fair use. The doctrine now not redistributes wealth from incumbents to the general public; it shifts wealth in the opposite direction, from the general public to powerful firms.”

This shifting of wealth from the general public to powerful firms has now come into crystal clear vision with the emergence of generative AI models. To permit fair use to power this theft from the general public and individual creators needs to be unthinkable. I do know that fair use and fairness aren’t the identical thing, and that fair use can sometimes be unfair to a specific creator in light of the broader perceived value of allowing a specific use. This, I submit, will not be such a case. Fairness and a careful & balanced approach to fair use are wholly aligned. This Fair Use Week, let’s all commit to a human-centered future, rooted in an appreciation of the role of consent in defining our universe. The choice is chilling — strip-mining of the soul, leaving only inhospitable conditions for the flowering of humanity.



Please enter your comment!
Please enter your name here