Last month, Meta said it would use generative AI in ad creation by the end of the year. CEO Mark Zuckerberg doubled down during last week’s earnings call, calling AI a “key theme” for the company.

Not long after, Google announced generative AI ad plans of its own in the form of tools that will “remix” ads based on client goals.

Neither company has released details of what these AI updates will look like, but the changes are big news for advertisers, agencies, and everyone in between.

Google versus Meta: Google has offered more details on its generative AI ad integration than Meta.

Google’s AI ads will be integrated into its existing Performance Max program, which already places ads algorithmically. AI could “remix” images and text to be personalized based on audiences or contexts. Copy may change slightly to target a given geography, for example.

Meta’s AI ad ambitions are a bit less clear, but sound more like a creative suite. CTO Andrew Bosworth told Nikkei Asia that clients would be able to “ask the AI, ‘Make images for my company that work for different audiences.’ And it can save a lot of time and money.” How these plans fit into Meta’s existing ad infrastructure is unclear.

What’s already available? Creative tools like Canva offer generative AI image and text generation built-in, which could be useful for advertisers. Google and Meta’s ambitions would allow advertisers to take advantage of generative AI integration without exiting the companies’ ecosystems.

What are the risks? The AI arms race has everyone scrambling to keep up, putting clients and consumers at risk.

Google’s initial Bard demonstration had a factual error, and generative AI is prone to hallucinations, where it confidently invents false information. Using that same AI to create advertisements could be a brand safety risk.

The same risks hold for Meta, which has had trust issues. With AI legal questions hanging in the air, brands using Meta’s tools may put themselves at risk for litigation. And the risks go deeper. “Advertisers would do well to remember that Meta has a history of inflating metrics for its own benefit,” said our analyst Max Willens. “Advertisers and agencies should be approaching this with extreme vigilance.”

What does this mean for agencies? Ad agencies will be some of the first impacted by native generative AI tools from Google and Meta. Changes in agency responsibility as a result of ad tech updates isn’t new, said Dhruv Grewal, PhD, professor of ecommerce and marketing at Babson College.

To stay relevant, agencies should become experts in using AI tools and analytics, focus on creating novel and authentic content where AI cannot, and position themselves to manage integrated marketing campaigns across platforms, Grewal said.

This was originally featured in the eMarketer Daily newsletter. For more marketing insights, statistics, and trends, subscribe here.