Expanded Access and Enhancements to Generative AI in Performance Max
AI advances have been delivering significant benefits for advertisers. It is assisting them in creating more impactful campaigns and producing creative assets at the necessary scale, speed, and variety to effectively engage consumers in relevant ways. Marketers and agency partners play a central role in the creative process, leveraging their in-depth knowledge of their customers and business to provide AI with the appropriate inputs for the development of meaningful, high-performing creative content.
In November, we introduced AI-powered asset generation and image editing in Performance Max. These tools enable the generation of more text and image assets for campaigns with just a few clicks. Asset generation is now being rolled out globally in English, with additional language support on the way. The rollout of image editing will be completed in the U.S. by March, followed by global expansion in English.
Furthermore, the capabilities of these asset generation tools have been enhanced with the integration of Gemini, Google’s most advanced AI model. Performance Max can now generate long headlines, with sitelink generation set to launch soon. These new features will leverage the sophisticated reasoning capabilities of the Gemini model for generating textual assets.
In the near future, when we update our image generation models to Imagen 2, which is tailored for Google Ads, users will be able to generate lifestyle imagery using Performance Max that showcases people in action. Similarly, the image editing capabilities will also include the ability to generate and add backgrounds featuring people. Additionally, for images that are performing well, users will have the option to generate similar alternatives to further expand their creative concepts. Advertisers will have full control and the ability to review all suggested images for their campaigns. Notably, the generation of images of named individuals, including celebrities and public figures, will not be permitted.
Moreover, advertisers are prohibited from using Performance Max or asset generation for creating content that promotes violence, misrepresentation, or any other content that goes against Google Ads policies and generative AI usage guidelines. We also utilize technologies such as watermarking with SynthID to enable the identification of AI-generated images. Imagen2 and SynthID are technologies developed by Google DeepMind.
We are committed to further enhancing these new generation capabilities for advertisers in adherence to our AI Principles and practices, ensuring responsible and innovative advancements.
