Google recently announced new requirements for political advertisements featuring synthetic or AI-generated content. Starting in November, these ads must prominently display clear and conspicuous disclosures to alert viewers to the presence of AI or synthetic material. The policy applies to all political ads on Google and YouTube, encompassing image, video, and audio content.
AI Raises Concerns As Campaign Season Approaches
This development comes as the 2024 US presidential election season gains momentum, and numerous countries worldwide gear up for their own major elections. Simultaneously, advancements in AIartificial intelligence have made it increasingly accessible for individuals to create convincing AI-generated text, audio, and video content. This accessibility raises concerns among digital information integrity experts, who worry that these AI tools could be exploited to spread election-related misinformation, posing challenges for social media platforms and regulators alike.
Addressing Misleading Content in Political Ads
AI-generated content has already made appearances in political advertising. For instance, a video shared by Florida Gov. Ron DeSantis' presidential campaign in June featured AI-generated images depicting former President Donald Trump embracing Dr. Anthony Fauci. These images, seemingly aimed at criticizing Trump, were interspersed with real photos of the two figures, accompanied by a text overlay stating, "real life Trump."
Similarly, the Republican National Committee (RNC) released a 30-second ad in April in response to President Joe Biden's campaign announcement, incorporating AI-generated imagery to depict a dystopian future in the event of the president's reelection. Although the RNC ad included a small on-screen disclaimer, "Built entirely with AI imagery," some viewers failed to notice it upon initial viewing.
Google's policy update seeks to address situations in which synthetic content might mislead users. For instance, it would require labeling an ad featuring synthetic content that makes it appear as if a person is saying or doing something they didn't say or do.
However, Google's disclosure policy will not extend to synthetic or altered content that is inconsequential to the ad's claims, such as minor image adjustments, color corrections, or background edits that do not create realistic depictions of actual events.
Additionally, a consortium of leading artificial intelligence companies, including Google, pledged in July to adopt voluntary commitments proposed by the Biden administration. As part of this agreement, these companies committed to developing technical mechanisms, like watermarks, to inform users when AI-generated content is involved.
The Federal Election Commission is also actively exploring regulatory approaches to address the use of AI in political advertisements, recognizing the significance of this issue in maintaining the integrity of the democratic processes.
