Google rolls out in-search AI text-to-image generator for select users
The new feature is set compete against current market leaders like OpenAI's DALL-E and Midjourney.
In an Oct. 12 announcement, Google announced that it allows some users to create AI-generated images through its main search bar.
The new feature allows users to submit a prompt and receive up to four image results that users can later refine with additional information.
The company said the feature is available to users who have opted into its AI-powered Search Generative Experience (SGE). Users will mainly access the feature through the search bar, but it will also be advertised within Google Images results.
Google stated that it will add limitations to prevent the creation of “harmful or misleading content,” adding that user-created images will have watermarks and metadata labels to show that they were generated by AI. Google said it will soon introduce a feature called “About This Image” for verifying AI-generated and traditional images alike.
The company additionally said that its generative search features will allow users to create draft documents — apparently separate from its similar Bard feature.
Image feature resembles other services.
Google’s AI image generation features will compete with numerous similar services, such as Midjourney, Stable Diffusion, Nightcafe, and DALL-E.
However, while many of those services offer free versions, most require users to purchase credits to upgrade to the platform’s full potential. Google’s announcement did not mention payment tiers or if the features will be free; regardless, introducing the technology to a search engine with such a large consumer base will make it accessible to more users.
The closest competitor at present is Microsoft’s Bing Image Creator. That service went live in March and was upgraded in October. It requires users to sign up for an account but is otherwise freely available through Bing Chat prompts.