Meta to require AI disclosure in political ads on Facebook and Instagram

Meta Platforms has announced that starting in 2024, advertisers on Facebook and Instagram will be obliged to reveal when they utilise artificial intelligence (AI) or other digital techniques to manipulate or create political, social, or election-related ads. This policy change, introduced by Meta, the second-largest digital advertising platform globally, aims to enhance transparency in online advertising.

Advertisers will be required to disclose whether their modified or newly created ads depict real individuals as saying or doing things they haven't, or if they create digitally-generated, lifelike individuals who do not exist. Moreover, advertisers will need to disclose if their ads depict events that never occurred, manipulate real event footage, or present real events without using the authentic visual, video, or audio recordings.

Meta's AI policy

These updates to Meta's policies follow their earlier decision to prohibit political advertisers from using generative AI ad tools. This decision came shortly after the company had announced plans to expand advertisers' access to AI-driven advertising tools capable of swiftly generating backgrounds, modifying images, and creating variations of ad copy based on text prompts.

Google, the largest digital advertising company under Alphabet, introduced similar image-customising generative AI ad tools last week. Google also pledged to exclude politics from its products by blocking a list of "political keywords" from being used as prompts for their AI tools.

In the United States, concerns amongst lawmakers have arisen regarding the use of AI to produce deceptive content in political ads, including deepfakes that falsely depict political candidates. The emergence of new "generative AI" tools has made it more affordable and accessible to create convincing deepfake content.

Meta has previously taken measures to prevent its user-facing Meta AI virtual assistant from generating photorealistic images of public figures. Nick Clegg, Meta's top policy executive, emphasised the need for rule updates in the realm of generative AI for political advertising.

Notably, the new policy will not mandate disclosures for digital content alterations deemed "inconsequential or immaterial to the claim, assertion, or issue raised in the ad," such as image resizing, image cropping, color correction, or image sharpening, as stated by Meta.

(With Reuters Inputs)