Patrick K. Lin

In the internet’s early days, First Amendment values laid the foundation for a free and open cyberspace. For a time, there appeared to be agreement among legislators, industry, and the public that the internet should be afforded great deference for the sake of innovation and technological development. In recent years, however, powerful technology companies have appropriated free speech principles to maintain a business model that prioritizes unfettered user engagement and data extraction without the threat of litigation. Consequently, laws rooted in the First Amendment, designed to redistribute legal rights from incumbents to the public now do the opposite: concentrate legal rights away from the public and to powerful companies. No industry has benefitted more from this shift than the technology industry. This Paper grapples with the current state of First Amendment interpretation, particularly as it relates to the technology industry, which has drifted from a decentralizing to a centralizing orientation. It offers reflections on the causes and consequences of interpretive missteps and proposes high-level principles for reversing this drift. To that end, this Paper surveys two examples: Section 230 of the Communications Decency Act of 1996 and fair use in the Copyright Act of 1976. Section 230 was passed in 1996 to nurture the internet in its infancy. However, early judicial interpretations of Section 230 failed to sufficiently consider changes in technology, entrenching virtually impenetrable liability protections for social media companies. Fair use, a redistributive legal doctrine that historically provided exceptions to educators, critics, and parodists, is now the go-to legal defense for AI companies to justify the theft of human-created work to train their models. These examples represent a shift in First Amendment interpretation that prioritizes powerful technology companies’ interests. Ultimately, this Paper concludes that a redistributive approach to First Amendment interpretation—one that explicitly accounts for power asymmetries and strives for distributive justice—could reverse speech’s ideological drift, and foster a safer and more democratic internet.

The fair use doctrine has been tech companies’ most valuable legal justification for systematically extracting and scraping content from the internet to train their generative artificial intelligence tools. Although fair use has historically protected the public from the monopolization of creative expression and intellectual property, in its current state, the doctrine enables powerful tech companies and movie studios alike to hoard copyrighted content while disenfranchising individual artists and creators. In May 2023, the United States Supreme Court ruled on Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith, a dispute about whether Andy Warhol’s use of Lynn Goldsmith’s original photo of Prince is a fair use. The Court concluded that a copy is not transformative under fair use when it is used commercially for a substantially similar purpose as the original. The decision affirms that the fair use inquiry must look beyond mere similarities and differences between a copy and the original. Instead, courts must grapple with how a copy is used. This Article examines the Supreme Court’s reinterpretation of the fair use doctrine in the Warhol decision in the context of creative expression and generative AI. More specifically, the majority’s emphasis on purpose and market competition tips the scale in favor of artists and creators concerned about generative AI models trained on copyrighted work. This Article also argues that the fair use analysis in Warhol recontextualizes Authors Guild v. Google, weakening a defense tech companies have relied on for ingesting copyrighted material to train AI models. Lastly, this Article considers the entertainment industry’s response to generative AI, including the WGA and SAG-AFTRA strikes, and proposes changes to the fair use doctrine to avoid a bleak future for art and creative expression. Ultimately, the Warhol ruling introduces some refreshing complexity back to the issue of fair use, particularly in context of generative AI and AI image generators. Still, many legal and existential questions remain. For instance, given the potential of AI to relieve society of dangerous and unfulfilling labor, why have tech companies opted to devalue labor that is abundant, meaningful, and undervalued?

The impact of artificial intelligence (AI) expands relentlessly despite well-documented examples of bias in AI systems, from facial recognition failing to differentiate between darker-skinned faces to hiring tools discriminating against female candidates. These biases can be introduced to AI systems in a variety of ways; however, a major source of bias is found in training datasets, the collection of images, text, audio, or information used to build and train AI systems. This Article first grapples with the pressure copyright law exerts on AI developers and researchers to use biased training data to build algorithms, focusing on the potential risk of copyright infringement. Second, it examines how the fair use doctrine, particularly its public benefit consideration, can be applied to AI systems and begin to address the algorithmic bias problem afflicting many of today’s systems. Ultimately, the Article concludes that the social utility and human rights benefits of diversifying AI training data justifies the fair use of copyrighted works.

While artificial intelligence (AI) has been a subject of great debate in spaces such as due process, discrimination, and privacy, an area that is lacking in legal scholarship is the technology’s environmental impact. AI promises to be a silver bullet in the increasingly urgent fight against climate change, yet it comes with a considerable cost to our planet. Current industry trends involve AI models being trained on increasingly larger datasets and training methodologies that prioritize brute-force over efficiency. Thus, as AI models increase in complexity and size, so too does the computing power—and energy—required to train and deploy them. Every stage of AI research and development, from training the model, storing its data, and deploying it in the real world, consumes the Earth’s resources. The interplay between AI and climate change is further complicated by the fact that AI is often lauded as an essential component of the new clean energy economy. If AI is meant to be a critical component of our new clean energy economy, its ever-increasing energy consumption must be addressed. By analyzing the processes and industry trends that cause AI to be a burden on the environment, this Article argues for mandating transparency around energy usage; empowering the newly formed National Artificial Intelligence Initiative Office to direct sustainable AI design; and pushing data centers to adopt clean energy. The Article then proposes a policy framework that not only minimizes AI’s carbon footprint but maximizes its potential to address key climate change concerns. For if we remain complacent, the very technology that could save our planet could very well be one of its greatest antagonists. This Article will be published in the Fordham Environmental Law Review, Volume XXXIV.

For more than two decades, police in the United States have used facial recognition to surveil civilians. Local police departments deploy facial recognition technology to identify protestors’ faces while federal law enforcement agencies quietly amass driver’s license and social media photos to build databases containing billions of faces. Yet, despite the widespread use of facial recognition in law enforcement, there are neither federal laws governing the deployment of this technology nor regulations settings standards with respect to its development. To make matters worse, the Fourth Amendment—intended to limit police power and enacted to protect against unreasonable searches—has struggled to rein in new surveillance technologies since its inception. This Article examines the Supreme Court’s Fourth Amendment jurisprudence leading up to Carpenter v. United States and suggests that the Court is reinterpreting the amendment for the digital age. Still, the too-slow expansion of privacy protections raises challenging questions about racial bias, the legitimacy of police power, and ethical issues in artificial intelligence design. This Article proposes the development of an algorithmic auditing and accountability market that not only sets standards for AI development and limitations on governmental use of facial recognition but encourages collaboration between public interest technologists and regulators. Beyond the necessary changes to the technological and legal landscape, the current system of policing must also be reevaluated if hard-won civil liberties are to endure.

“When today’s technology relies on yesterday’s data, it will simply mirror our past mistakes and biases.” AI and other high-tech tools embed and reinforce America’s history of prejudice and exclusion — even when they are used with the best intentions. Patrick K. Lin’s Machine See, Machine Do: How Technology Mirrors Bias in Our Criminal Justice System takes a deep and thorough look into the use of technology in the criminal justice system, and investigates the instances of coded bias present at every level. In this book, you’ll learn how algorithms and high-tech tools are used in unexpected ways: suggesting which neighborhoods to police, predicting whether someone is more or less likely to commit a crime, and determining how long someone’s prison sentence should be. Machine See, Machine Do takes you on an eye-opening journey of discovery, encouraging you to think twice about our current system of justice and the technology that supposedly makes it more “objective” and “fair.” If you are someone who cares deeply about criminal justice reform, is curious about the role of technology in our day-to-day lives, and ultimately believes we should aspire to make both of these spaces more ethical and safe, this book is for you.