Scarlett Johansson and the AI Voice Controversy

May 20, 2024
Scarlett Johansson and the AI Voice Controversy: Did OpenAI Copy Her Voice?

Hollywood actress Scarlett Johansson raised eyebrows recently. She accused OpenAI, the company behind ChatGPT, of copying her voice for their AI assistant, Sky. Johansson claimed Sky's voice sounded "Very similar" to hers. This sparked a debate about AI development and celebrity voices.


Scarlett Johansson reportedly felt "shocked" and "angered" after hearing Sky's voice. She had previously declined a request from OpenAI CEO Sam Altman to license her voice for their project. This led her to believe they had used her voice without permission.

OpenAI, however, strongly denied the accusation. They insisted they never intended to imitate Johansson's voice. OpenAI stated, "We believe AI voices should not deliberately mimic a celebrity's voice." They claimed Sky's voice belonged to a different actress.

Mr. Altman said the voice of the sky is not Scarlet Johannsson's and was never intended to resemble hers. We cast the voice actor behind Sky's voice before any outreach to Miss Y Hansen. Out of respect for Miss Y Hansen, we have paused using Sky's voice in our products.

The actress is the latest celebrity to question how individual rights can remain protected in a world where artificial intelligence and deep fakes are becoming even more pressing threats.

Scarlett Johansson and the AI Voice Controversy

The Washington Post investigated the controversy. They found documents suggesting OpenAI cast a voice actor for Sky months before contacting Johansson. Interviews with people involved in the hiring process revealed OpenAI wasn't seeking a Johansson sound-alike.

So who's right? Johansson or OpenAI? The truth may be unclear. Even if not an exact copy, it's possible Sky's voice shares similar qualities to Johansson's. Human voices naturally vary, but some share characteristics. Perhaps Sky's voice simply falls within that range.

Should we be concerned with how human they're making it? That's a great question. I mean, the number one rule in AI is to not try and anthropomorphize like what you're dealing with, and yet researchers consistently kind of fall into the trap around it.

Because of the way that AI is very different, yeah, it's different, but that line between humans and AI I see a red apple on a plate in the center of the table that is starting to fade. It took us, you know, 10 to 12 years of experiencing the harms of social media before we thought. Oh, maybe we should regulate this. I think we're moving faster now with AI, but we have to keep up. Perhaps AI could walk us through it and then exhale slowly.

AI Identity Theft?

This incident highlights the growing complexity of AI development. As AI becomes more sophisticated, the ability to mimic human voices raises ethical questions. Should celebrities have control over the voice data used in AI? Who owns the rights to a voice generated by AI?

We've seen AI and voices trying to impersonate other people in scams and things like that. Already Mike Prospero is the editor-in-chief of the tech website, Tom's Guide, and right now he says what the sad reality is. There is no simple solution as AI gets more sophisticated; it's going to be a lot easier to replicate people living or dead, and you know how that is used ethically, which is going to be an issue that everyone's going to have to grapple with.

The Johansson-OpenAI case doesn't have a definitive answer yet. But it has sparked important discussions about AI ethics and intellectual property in the digital age. As AI technology advances, these questions will likely become even more pressing.
Piyush Verma

Piyush Verma

I am a Blogger and Web Designer.

Next Post Previous Post