A recent AdAge article discussed the state of advertising inside ChatGPT and generative AI (genAI) more broadly.
Agencies see signs of maturity in the new ad business, such as StackAdapt joining, but it is still an experimental media buy for brands.
While 63% of U.S. adults say that ads alongside generative AI outputs would make them trust those outputs less, early performance data from ChatGPT ads paints a more optimistic picture for ads in genAI environments.
Engagement rates ads on ChatGPT are reportedly competitive with high-intent brand search, and brands are reaching incremental audiences that legacy search formats don't capture. Despite constrained inventory, a still-developing auction infrastructure, and CPMs that have swung wildly from $60 to as low as $15, performance metrics suggest that ads in genAI environments is a legitimate medium for which advertisers will need a strategy.
As genAI advertising evolves from an experiment into a substantial budget line item, platforms will need to build trust with advertisers as well as with consumers, and the industry can't afford to treat this ad format as a blank slate. The same questions that took years of industry work to address in legacy digital advertising environments don't disappear just because the content on our screens looks different or because these ads perform well.
Where do ads appear relative to model outputs?
How is brand suitability defined and enforced in a genAI environment?
What verification and measurement options are available for advertisers?
The best time for the industry to ensure accountability, transparency, and safety are integrated from the outset without hindering innovation is now, as the pipes are being built.