Deepfake Protection Act – Bill 001
Status – Proposed
Category – Technology & Artificial Intelligence
Goal – Protection of individuals from harmful AI-generated deepfakes while preserving creative expression
Ⅰ. Introduction
Taylor Swift, an iconic, world-renowned pop star, has dominated headlines in early 2024 not because of her hit song “Cruel Summer,” but because of explicit deepfake images of her circulating on the internet. These pictures were taken down quickly, but long after the damage was done, they were viewed and shared by millions of people. This incident quickly turned into an internet frenzy and had everyone discussing the regulations of artificial intelligence. If one of the most recognizable people in the world could become a target for false AI-generated deepfakes, what protections are there for everyone else? This exact question sparked my curiosity about proposing a Deepfake Protection Act.
Ⅱ. Who is affected, and why now?
Who is affected?
Anyone can become the victim of misused AI-generated deepfakes. Students, businesses, and public figures are all vulnerable, past President Obama being no exception. It is easy to make a deepfake, post it, and take it down, but the damage is irreversible.
Existing Laws
Under current laws, it is not possible to effectively limit the power or abuse of AI. These laws are already bypassed in many situations, making it hard to acknowledge or prove the involvement of artificial intelligence.
Fraud Law – When an individual purposefully deceives someone to gain money or anything of value.
Problem – This specific law has been involved in many cases that have used AI, but fraudulent law doesn’t focus on the involvement of AI-generated content. If a scammer has used AI, it is the scammer being punished; the company that made the AI won’t receive any penalty.
Privacy Law – Protection for individuals from AI-generated content shared about them without their consent, such as the Taylor Swift incident.
Problem – These laws were made before AI even spiraled. With deepfakes becoming more realistic, the Privacy Law does not always clearly address images that were fabricated rather than real.
We have laws for many situations, but just not ones that limit people who generate harmful content and post it within a matter of seconds. My bill proposal will address the rules for this specific topic and clear any gaps created by this newly advanced technology.
Ⅲ. What My Bill Would Do
I propose and advocate for AI disclosures. Any content generated or using AI should have a clear disclosure. This would allow all viewers of this content to be more knowledgeable about the content they are seeing, and it would allow them to make more informed decisions without the restriction of their creativity.
Many victims of deepfakes often struggle with the breadcrumbs that any harmful AI-generated content leaves. As part of my bill, I would like to propose a clearer legal process to report harmful AI-generated content when deepfakes cause reputational, financial, and, most importantly, emotional harm. I propose a process of requesting its permanent removal and seek legal remedies for the problem.
In my bill, I want to propose AI-specific courts. If we were to take a murder case, it would take months, maybe even a year to finish the case, and the damage is already done. For a deepfake or artificial intelligence case, the damage can quadruple by the day if the judgment is delayed, and the case must be attended to quickly. For that reason, I urge to have courts specific for these types of cases.
Ⅳ. Possible Concerns & Challenges
Every good bill has its challenges. The main challenge I think my proposed bill would face is the implementation of free speech and the fact that it shouldn’t be restricted. Just as AI evolves, this bill can learn with it. The bill is adaptable in growing, and lawmakers must ensure that this bill doesn’t restrict freedom of expression or creativity. In response to all the misconceptions about AI, it is a greatly positive and effective tool. My bill would propose to protect all the creative, educational, and research uses of AI-generated content. This content should not be used to deceive or harm other individuals.
Ⅴ. Conclusion
This bill’s whole purpose is to protect individuals against the dangers and harm of deepfakes and other generated content. Like every bill, it isn’t perfect; it may not address every single challenge out there, but it is a bigger step towards protecting our society and using a powerful tool to protect everyone. At the end of the day, our society needs to acknowledge the “elephant in the room”. If an automatic car caused a crash, do you blame the car or the company that made the car? Laws should require an “opt-in” rather than an “opt-out”; the only reason it isn’t set that way is that big companies such as Meta, OpenAI, and Google, and many other technology companies, manipulate politicians so laws are biased in their favor.