SAG-AFTRA on AI Taylor Swift deepfakes, George Carlin special


The Screen Actors Guild – American Federation of Television and Radio Artists (SAG-AFTRA) put out a statement on Jan. 26, 2024, hitting back at two AI scandals in recent weeks: explicit AI deepfakes of musician Taylor Swift circulating on X (formerly Twitter) and the wider internet, and a comedy special YouTube video that impersonates the late comedian George Carlin and was marketed as AI generated, but which a spokesperson of the creators said was actually human written and an elaborate joke.

It makes perfect sense and is in keeping with the union’s prior position of skepticism toward AI during its strike and ultimately, agreement to renew production with studios, with the agreement that human actors would control their own AI likenesses generated on film and TV productions.

A scandalous month for AI generated content

It’s unclear if Taylor Swift is a SAG-AFTRA member, but she’s acted before and is rumored to be in the upcoming Deadpool 3, and during the past week, various X accounts and websites such as Celeb Jihad posted deepfaked photorealistic images of her in various lewd and explicit scenes involving fans of her pro American football player boyfriend Travis Kelce’s NFL team the Kansas City Chiefs.

X was slow to stop the spread of the images (which remained up for 17 hours and received 45 million views for the most viewed posts), but did finally begin removing them and associated accounts, and today blocked all searches of Swift’s name.

Meanwhile Carlin was made the subject of a new comedy special titled “George Carlin: I am Glad I’m Dead,” released on January 11, 2024, by “Dudesy” a so-called comedy AI program created by comedian Will Sasso formerly of “Mad TV” and podcaster Chad Kultgen, to much criticism online by fans and ultimately, a copyright infringement lawsuit from Carlin’s daughter. The video was taken offline after the lawsuit was filed.

“The development and dissemination of fake images — especially those of a lewd nature — without someone’s consent must be made illegal. As a society, we have it in our power to control these technologies, but we must act now before it is too late,” SAG stated.

Early on, some alleged the two are performing an elaborate bit as an AI program to spoof the hype and fears around the tech, possibly using AI tools to help them with this, but a representative for the duo told Ars Technica in article published today that “It’s a fictional podcast character created by two human beings, Will Sasso and Chad Kultgen…The YouTube video ‘I’m Glad I’m Dead’ was completely written by Chad Kultgen.”

But just yesterday, SAG-AFTRA released a statement under the impression the Carlin special was AI-generated, alongside its concern for the real AI-generated Swift explicit deepfakes. Read it in full here:

January 26, 2024 

SAG-AFTRA is working with lawmakers who understand the importance of acting immediately to pass legislation that protects individuals’ voices and likenesses — both in the case of living and deceased performers.

The sexually explicit, A.I.-generated images depicting Taylor Swift are upsetting, harmful, and deeply concerning. The development and dissemination of fake images — especially those of a lewd nature — without someone’s consent must be made illegal. As a society, we have it in our power to control these technologies, but we must act now before it is too late. SAG-AFTRA continues to support legislation by U.S. Rep. Joe Morelle, the Preventing Deepfakes of Intimate Images Act, to make sure we stop exploitation of this nature from happening again. We support Taylor and women everywhere who are the victims of this kind of theft of their privacy and right to autonomy.

Similarly, George Carlin’s legacy was formed out of a lifetime of lived experience and decades of honing his astute observational skills while demonstrating a tireless dedication to the art of comedy. Companies or individuals attempting to replicate an individual’s creative work for profit, publicity or clout are infringing upon intellectual property rights and devaluing the essence of human artistry. We must remember that everything generated by A.I. originates from a human creative source and human creativity must be protected. SAG-AFTRA’s legal team has been working with lawmakers who understand the importance of acting immediately to pass legislation that protects individuals’ voices and likenesses — both in the case of living and deceased performers. Families should not have to see their loved ones exploited for profit. We support the Carlin family, who are pursuing legal action against those using the comic’s work without seeking consent or offering compensation. 

WH backs Congressional action, and two Ai image bills have been introduced

It’s not just SAG-AFTRA calling for laws to change in the wake of the rapid spread of AI imagery from generator models and programs such as OpenAI’s DALL-E 3 (which powers Microsoft Designer and Bing Image Creator), open-source model Stable Diffusion (made by Stability AI and Runway), and Midjourney: the White House spokesperson said this about the Swift illicit AI deepfakes in a press conference:

“We are alarmed by the reports of the…circulation of images that you just laid out – of false images to be more exact, and it is alarming,” White House Press Secretary Karine Jean-Pierre told ABC News White House Correspondent Karen L. Travers.

“While social media companies make their own independent decisions about content management, we believe they have an important role to play in enforcing their own rules to prevent the spread of misinformation, and non-consensual, intimate imagery of real people,” she added.

She also stated: “Of course Congress should take legislative action. That’s how you deal with some of these issues,” according to The Verge.

Republican Congressman Tom Keane of New Jersey has introduced a bill calling for all AI generated images to be labeled, and joined with his House of Representatives colleague, Democratic Congressman Joseph Morelle of New York, on the Preventing Deepfakes of Intimate Images Act, which would “make non-consensual deepfakes a federal crime,” and allow purveyors to be fined and sent to federal prison for 10 years.

Both bills are awaiting voting through committee and onto the full House and a Senate version being introduced, before they can be voted on by the full Congress and signed or vetoed by the President. But at this rate, with seemingly most people up in arms about the issue, there seems a high public demand for rapid action from lawmakers. The big questions: will they act, when, and how?

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.





Source link