AI’s Impact: Taylor Swift Endorses Harris After Election Influence

• Experts worried AI might disrupt elections globally.
• Contrary to fears, AI has not significantly affected the 2024 election cycle.
• Instead, it has primarily generated memes and influenced pop culture endorsements.

For some time now, experts have expressed concerns that artificial intelligence could mislead voters and cause chaos in elections all over the world.

However, the widespread disruptions that many anticipated have not materialized. Instead of the feared deepfakes and misinformation campaigns, AI is mostly being utilized for humorous meme creations.

One of the most notable outcomes of AI’s involvement this election cycle has been its role in encouraging Taylor Swift to publicly endorse Kamala Harris for president.

In a recent Instagram post, Swift revealed that her endorsement was partly inspired by an AI-generated image of herself shared by Donald Trump. The image portrayed her in an oversized American flag hat with the phrase “Taylor Wants You To Vote For Donald Trump.”

“This image really brought to light my concerns about AI and the dangers of spreading misinformation,” Swift stated in her post. “It made me realize the importance of being clear about my intentions for this election as a voter. The most direct way to combat misinformation is through truth.”

Swift’s apprehensions about AI are echoed by experts and media outlets, who have warned that AI could lead to a “tech-enabled Armageddon,” suggesting we’ve only glimpsed the potential dangers as AI continues to evolve.

Despite attempts to mislead voters using AI, such as a phony robocall featuring Joe Biden or a deepfaked video of Kamala Harris, these efforts have not effectively deceived many individuals.

Various AI creations have remained largely apparent as memes and satirical clips shared across social media. Fact-checking initiatives, including those on platforms like X, have swiftly debunked any AI content that seems convincing.

Moreover, while some foreign entities may have harnessed AI to disseminate false information, the outcomes appear to have been exaggerated.

According to Meta’s latest Adversarial Threat Report, even though Russian, Chinese, and Iranian disinformation operations have integrated AI into their strategies, the overall effectiveness and efficiency of these “GenAI-powered tactics” have remained limited.

Similarly, Microsoft released a Threat Intelligence Report in August that downplayed the effectiveness of AI in foreign influence campaigns. They identified several influence operations from Russia and China, stating that these entities have employed generative AI but with little to no actual impact.

Microsoft further noted that numerous actors have sought to incorporate AI into their endeavors, but many have reverted to previously successful strategies involving simple digital alterations and trusted labels affixed to misleading information.

It’s worth mentioning that the trend isn’t isolated to the United States. Recent elections worldwide also did not experience significant disruptions from AI.

An analysis conducted by the Australian Strategic Policy Institute found that in reviewing AI-generated misinformation related to the UK’s July elections, there was no evidence of the predicted “tsunami of AI fakes” aimed at political figures.

Sam Stockwell, a researcher with ASPI, pointed out that only a handful of AI-generated content examples went viral during the campaign season. However, while these instances did not sway a significant number of voters, they did result in an increase in online harassment targeted at individuals depicted in the fakes, along with confusion among the audience regarding the authenticity of the content.

A study by the UK’s Alan Turing Institute highlighted that, of 112 national elections occurring or scheduled since the start of 2023, only 19 experienced any form of AI interference.

The report concluded that instances of AI misuse during elections are rare and often amplified by mainstream media coverage, amplifying public fears and potentially overstating the real threat AI poses to electoral integrity.

Despite researchers concluding that the current influence of AI on specific election outcomes is limited, they acknowledged that the potential dangers are indicative of broader systemic threats to democratic processes.