Discover more from The Ethical Tech Project

We are a think and do tank that seeks to build, design, align and implement a blueprint for the ethical use of data on the internet.

Hollywood Won’t Be The Last Industry To Fight Over AI Rights

Writers and actors are just the canaries in the coal mine. Maintaining agency over one's data will be even more important in a world where AI has unfettered access to public information

The Ethical Tech Project

Sep 6, 2023

From Hollywood to Corporate Training — We Can All Be Duplicated

After covering privacy last week, this week we’re looking at the second of our 5 principles: Agency.

Agency: To preserve data dignity, people should be given choice and control over how their data is used and have the power to change their decision at any time – and those choices should be forced across all the data systems that process that data.

Our guiding belief is that people should be given a choice and control over the use of their data and be able to change their decisions wherever their data is used. How to actually implement that principle into practice, however, is a live and messy debate.

(Source: Getty Images)

This battle is a big one. Right now, the fight for agency is being fought in the boardrooms of Hollywood, where striking writers and actors are in a prolonged standoff with studios over AI policy. Agreeing on standards over the use of AI is being called the “wild card” in the strike negotiations with no clear resolution in sight.

Writers want guarantees AI won’t be used to write or rewrite scripts, used for source material, or trained on past scripts. Actors want guarantees that the likeness of film extras won’t be scanned (which was already happening) and used in perpetuity without compensation. The studios have offered partial concessions, but neither side can agree and continue to debate each other. An episode from the latest season of Black Mirror, “Joan is Awful,” neatly anticipated the ethical unease over a world where computers generate content based on personal data and likenesses. In that episode, the individuals being portrayed have no legal method for stopping damaging portrayals of themselves after signing away the rights to their likeness when they agreed to a streaming platform’s terms and service agreement.

These AI debates aren’t just a labor spat confined to Tinseltown. They are the first of many debates over how people’s data (including yours) can be used. AI startups that create digital clones of people for purposes like corporate training videos have already reached unicorn valuations. Companies like Zoom are backpedaling after facing backlash for updating their terms to allow the use of customer data to train AI models. Who knows, that last all-hands meeting where you forgot to mute your mic before sneezing could be used to train future models.

Is the Rock’s face as an easter island head delightful AI art? Absolutely. Does the Rock have any legitimate claims as to how the image is used? Unclear! (source)

The big ethical question is: Where should we draw the line to create an AI-powered future that preserves agency over one’s data?

Agency implies people should have meaningful control over how their likeness is licensed and used — but what does that really mean?

Does this mean I can fully sign away my likeness in perpetuity for all use cases?
Is a person’s image their own intellectual property even if they post it publicly?
What about if a face is simply one of millions used to create human-like avatars in the future?
Does that person have any meaningful claim to the commercial use of how that avatar is used?

Any legal regime over likeness also must consider how to navigate unethical companies that operate outside the law. Even the most well-defined set of laws and norms around likeness will be disregarded by unscrupulous or law-breaking organizations. We’re already seeing the horrific rise of uses like nonconsensual deepfake porn trained on specific people’s likenesses collected from their public social media posts. How should responsibility be divided between government and industry to eliminate harmful uses and ensure people’s claims to data agency is protected? Do we even have the right legal and technical tools to make sure harmful uses of likeness are identified, stopped, and punished?

This battle is a messy one, and we are only in the opening skirmishes in the battle over likeness rights. Instead of being rushed into a world of half-baked policies, the public deserves a thoughtful approach from companies and public leaders that prioritizes individuals’ rights and prioritizes human agency.

Tell us your thoughts: What rights and protections should individuals have over their likeness? Do effective tools and laws exist, or does this require a totally new legal framework?