Taylor Swift's Harris endorsement, AI 'fears' likely to spark dialogue around deepfakes

“Recently I was made aware that AI of ‘me’ falsely endorsing Donald Trump’s presidential run was posted to his site,” Swift wrote in her Instagram post. “It really conjured up my fears around AI, and the dangers of spreading misinformation. It brought me to the conclusion that I need to be very transparent about my actual plans for this election as a voter. The simplest way to combat misinformation is with the truth.”

Sexually explicit deepfakes of Swift also went viral online in January.

Swift’s level of fame is such that her concerns about AI could have a measurable impact on the political discourse surrounding the technology, according to Henry Ajder, an expert on AI and synthetic content. “It will not be lost on politicians that this was something she cited in her statement,” he says.

And with 283 million Instagram followers, her concerns around AI can “quickly bypass a marketer’s biggest hurdle: capturing attention,” says Laurel Cook, associate professor of marketing who has researched the social implications of deepfakes.

A number of other A-listers have also publicly spoken out against the use of deepfakes. In May, Scarlett Johanssson threatened legal action against OpenAI after the company released an AI model with a voice which, she claimed, was nonconsensually modeled on her own. (OpenAI denied the claim but removed the voice “out of respect for Ms Johanssen,” its CEO Sam Altman wrote in a statement.) And last month, Tom Hanks warned his followers on Instagram about AI-generated ads that depicted him as endorsing, in his words, “miracle cures and wonder drugs.”

Such high-profile complaints should be heeded by brands that are keen to deploy generative AI for marketing purposes, according to Kristy Roschke, director of the News Co/Lab at the Walter Cronkite School of Journalism and Mass Communication at Arizona State University. “People like Taylor Swift [and Johannson] have the ardent fanbase and recognition to shape norms and expectations that favor protecting their likenesses,” she says. “These early examples are strong warnings to brand marketers, who may be enamored by the novelty and ease of using generative AI, that they should not use it in this way.”

The rise of AI-generated deepfakes has caused excitement for many working across creative industries, like marketing and entertainment. AI makes it possible to recreate celebrity likenesses cheaply and quickly, and brands can even (virtually) bring beloved icons back from the dead to give their marketing a little more oomph.

But such emerging use-cases have also led to fears about the possibility for the tech to be misused – by, for example, using a celebrity’s AI-generated likeness without their consent or to advance some political or ideological agenda.

The current lack of regulation around the use of generative AI in the United States means that the court of public opinion has been a major force in determining which uses of the technology should be considered ethical or unethical. Fears around the tech’s potential to impact elections were heightened in January after an AI-generated robocall mimicking the voice of President Biden was sent to voters in New Hampshire. And during his recent feud with Kendrick Lamar, Drake’s use of AI to recreate the voice of Tupac Shakur in a song drew sharp criticism.

Some marketing experts, meanwhile, have voiced concerns about the potential for generative AI tools to perpetuate harmful biases – in their depiction of women, for example – or, somewhat counterintuitively, to contract rather than expand the industry’s creative horizons.

Harris and Trump have differing stances toward the technology and its growing influence in popular culture. Trump seems to regard Biden’s AI Act as being antithetical to domestic industry at a time when adversaries like China and Russia are rapidly advancing in their own technological capabilities; he’s vowed to toss the Act if he’s reelected on the Republican ticket in November.

Harris, born just outside Silicon Valley and the former attorney general of California, has spent much of her vice presidency deepening her ties with big tech leaders and has repeatedly stated her position that government should play an active role in guiding the evolution of the AI industry.

Harris’s home state – helmed by a fellow Democrat, Governor Gavin Newsom – is considering a landmark bill that would require tech companies and social media platforms to ensure that AI-generated content is identifiable as such. Another California bill, currently awaiting Newsom’s approval (or veto), would require tech giants to adhere to certain safety standards when deploying powerful AI models.

The deepfakes of Swift and other notable public figures that have recently gone viral are only a taste of what’s to come, given how rapidly the technology is advancing, argues Ajder, the synthetic media expert. “We are going to see increasingly sophisticated capabilities from AI tools and tools [allowing for] a level of realism … that will make the examples we are seeing today look quaint,” he says. “The future is a space where the synthetic and the authentic increasingly are blurred.”

For more on the latest happenings in AI and other cutting-edge technologies, sign up for The Emerging Tech Briefing newsletter.