Google requires political advertisements to indicate whether they are generated by artificial intelligence.

2023-09-08

Starting this fall, AI-generated political ads appearing on the Google platform must disclose whether they contain "synthetic" content to users.

As concerns grow about AI generating false ads and other misleading content that can harm people, Google has taken a step towards controlling the damage. In an update to its political ad policy, Google will require advertisers to explicitly state when their ads include synthetic content created by AI. This policy specifically targets political ads that "depict people or events truthfully or realistically."

The policy will go into effect in mid-November, and it stipulates that explanations regarding the use of AI must be clear and prominently displayed where users are likely to notice. This requirement applies to political ads with images, videos, or audio that Google provides on its own platforms (such as YouTube) and as part of the display network of third-party websites.

However, this policy does not include ads that have undergone simple editing. In other words, as long as the ads do not fabricate real individuals or events, adjustments such as resizing, cropping, color correction, error correction, and even background removal are allowed for images or videos.

Google provides several examples that require clarification. One is political ads with AI-generated content that make it appear as if individuals are saying or doing things they haven't. Another example is AI-generated ads that alter the perspective of actual events or depict scenes of real events that never occurred.

Distorting or manipulating the truth in political ads to make opponents look bad has a long history. However, the social media and AI era have exacerbated this problem.

Using AI, political campaign teams or other parties can easily create seemingly authentic false images, videos, or audio clips that portray candidates saying or doing things they haven't. Biased voters may take these lies as truth without fact-checking.

In response to concerns about political ads, Google has previously taken other measures. In 2018, the company began requiring advertisers of all political ads to disclose who paid for the ads to verify their identity.

Google's transparency report includes political ads from the United States and other countries to help people understand who purchased specific ads, how much money was spent, and how many times the ads were viewed. In 2019, the company expanded transparency to include ads related to state-level candidates, political parties, and voting initiatives.

Google's policy also prohibits the use of deepfakes and other false content that can deceive people on political and social issues. The company uses automated systems and manual reviews to identify and remove ads that violate its policies. In 2022, Google removed 5.2 billion policy-violating ads and blocked 2.6 million election ads that failed to complete the verification process.

A Google spokesperson stated, "Over the years, we have provided users with a high level of transparency for election ads, including 'paid for' disclosures and a public ad library that provides more information about the election ads they see on our platform."

The spokesperson added, "Given the increasing prevalence of tools for creating synthetic content, we are further expanding our policy to require advertisers to specify which part of their election ads contains digitally altered or generated materials. This will help further ensure the authenticity of political ads and provide voters with the information they need to make informed decisions."

RECENT AI NEWS

RECENT AI TOOLS