The Benefits of Well-Regulated AI

Last week, President Biden issued a landmark Executive Order on artificial intelligence that will shape the future of AI in the United States and globally. This action showed great foresight on part of the Biden administration, as with each passing day, AI is only becoming more prevalent in almost every industry in the country.

It is crucial to set the stage now, with regulations and guardrails, to protect Americans from AI fraud and algorithmic discrimination — not only for the future of AI in general, but particularly for the entertainment industry.

As a VC firm that straddles the world between Hollywood and Silicon Valley, the regulation of AI is directly relevant to our interests and pursuits. We are already seeing AI begin to pervade the world of TV and film. For the time being, it’s often used for special effects purposes, but it’s quickly on the verge of being used for screenwriting, and some want to eventually use it for the wholesale creation of characters and long form content.

The tea leaves are evident to anyone in the industry; it was no surprise that the WGA and SAG-AFTRA made sure to include AI provisions into their recently negotiated contracts with the studios. Along with hope about the democratizing power of AI, there is also a real worry that AI produced content will not only replace work done by creators, but also be used to deceive audiences and wreak havoc and distrust. To help protect against this, the Biden administration included the following provision in the Executive Order:

“Protect Americans from AI-enabled fraud and deception by establishing standards and best practices for detecting AI-generated content and authenticating official content. The Department of Commerce will develop guidance for content authentication and watermarking to clearly label AI-generated content. Federal agencies will use these tools to make it easy for Americans to know that the communications they receive from their government are authentic — and set an example for the private sector and governments around the world.”

Discerning between AI-generated content and authentic content is a solid first step in setting clear boundaries. There is nothing wrong with the use of AI, but people should know when they are consuming content produced by AI so they have the choice on what type of content they want to support and consume.

Another reason it is important to know if content is AI-produced is because of the threats of deepfakes. A deepfake is a manipulated video or audio that uses AI to create a convincing and often deceptive portrayal of someone saying or doing something they never did. Deepfakes are dangerous because they can spread false information, impersonate individuals, manipulate public opinion, and potentially incite harm or damage reputations.

Although the Executive Order only requires the government to watermark AI-generated content, they are setting a precedent that others should follow. If all AI material is distinguished this way, it can help Americans learn what AI looks like and help them to spot bad actors who are using deepfakes.

Another concern with the use of AI in content creation is that it may further silence people of color and replace them their stories with AI-generated material. Hollywood has historically overlooked creators of color; and with AI, content creators may find it is easier to create a Black character digitally than it is to cast an actor and include them in the storytelling narrative.

It’s why the Executive Order also included a section on advancing equity and civil rights, with one provision, “directing agencies to combat algorithmic discrimination.” Some people have scoffed at this, but the algorithms that govern AI are only as good as the folks who program them. We have seen over and over again that those creating AI technology often forget to consider diversity when making their algorithm.

Overall, I am glad to see the Biden Administration taking AI seriously and working to set up guidelines. Whether one likes it or not, AI is here to stay. If we can set clear guard rails from the start, as AI accelerates, we can ensure that AI is a tool that enhances our lives and democratizes creation–rather than becoming a dangerous tool that misleads, and only further excludes diverse voices from breaking out of the fold.