Deepfakes: The Next Brand Safety Challenge

Deepfakes: The Next Brand Safety Challenge

PLUS: Reddit's new Ads API... LinkedIn's new video ads... Changes to Instagram carousels... and more.

Keeping Your Ads Away From Deepfakes

A lot of brand safety tools for advertisers deal with adjacency — how close your ads get to pro-Nazi content, for instance. Your brand might not want its ads for school backpacks to be directly below a post about silencers on AR-15s.

Until now, the list of things to stay away from have been topic-driven: Violence, war, gambling, and so on. But now, we’re starting to see protection based on the origin of the content.

Integral Ad Science says it’s working on a new brand safety tool designed to keep deepfake content away from ad content.

This, of course, happening as AI-generated misinformation and fake images are continue to grow. A recent BBC investigation found that TikTok’s algorithm was recommending fake videos of political candidates making controversial statements.

IAS says its deepfake tool would be an industry first, though OpenAI has said it’s working on one of its own — both companies clearly trying to capitalize on brand safety anxieties.

Most analysts say the issue of deepfake ad adjacency is likely to worsen in coming years as generative AI becomes more widely available.