In a recent update to its political content policy, Google announced that it will implement new regulations mandating transparent disclosure regarding the use of artificial intelligence (AI) in political advertisements. As of November, any synthetically altered images or sounds in such ads must be accompanied by an unmistakable notification located in an area users are likely to notice.
The new rules represent Google’s response to the increasing ubiquity and sophistication of AI in political advertising. While the deployment of manufactured images, audio, and video content in political ads is not a novel phenomenon, the advent of AI generators has rendered the fabrication process not only more effortless but also more convincing. Even at this nascent stage of the 2024 race, several presidential campaigns, including that of Florida’s GOP Governor Ron DeSantis, are utilising these contemporary tools.
A noteworthy instance was the release of an entirely AI-curated advertisement by the Republican National Committee in April. This ad depicted an apocalyptic vision of America’s future under a second term of President Joe Biden, with photorealistic images of desolate, boarded-up storefronts, martial law on the streets, and terror-stricken citizens amidst an immigrant influx.
Moreover, in June, DeSantis’ campaign publicised an attack ad featuring AI-generated images of his GOP primary contender, former President Donald Trump, in a friendly embrace with infectious disease expert Dr. Anthony Fauci.
The Federal Election Commission, in light of such advancements, has initiated a process to examine the potential regulation of AI-crafted ‘deepfakes’ in the context of political advertising ahead of the 2024 election. The urgency of this issue has also captured the attention of Congress, leading to discussions about instituting legislative safeguards against deceptive AI-created content. Senate Majority Leader Chuck Schumer and other lawmakers have vocalised their intentions to implement such measures.
Legislation relating to the use and potential misuse of deepfake technology has also been debated or enacted in several states.
However, Google’s policy update does not categorically proscribe the use of AI in political advertising. It allows for exceptions where the AI-induced alterations or generation are deemed irrelevant to the advert’s claims. This encompasses editing techniques such as image resizing, cropping, colour correction, defect rectification, or background modification, provided they do not falsify actual events.
