Meta Enhances AI Image Generation for Ads

135

Meta Platforms, the parent company of Facebook and Instagram, has announced plans to broaden its range of generative AI ad products. This expansion will introduce tools capable of automatically producing various image variations and overlaying text onto them.

During a press conference, Meta executives disclosed that the tool would initially be launched in a test format without the watermarks typically applied to all images generated by Meta’s user-facing AI assistant. John Hegeman, Meta’s head of monetization, explained that the company is still refining how labeling will function for ads and intends to share guidelines before the tool’s global rollout, anticipated around the end of this year.

This move comes as Meta invests significant resources into developing and supporting its generative AI models while aiming to persuade advertisers of the benefits of automating creative aspects of their campaigns. In a parallel development, Google recently announced a comparable expansion of its AI ads tools in February. Google’s ads created through these tools will be labeled using SynthID watermarking technology, developed by its AI research lab, DeepMind.

Meta’s image generation tool will enable advertisers to upload product images and generate additional versions by adjusting product orientation or depicting people using them in different settings. Furthermore, Meta is enhancing its text-generation capabilities for headlines and key selling points, along with the ability to directly overlay text onto generated images.