In May 2022, Google unveiled an AI image generator known as "Imagen," which creates hyper-realistic pictures based on descriptive texts — for instance, a Persian cat playing guitar or a panda riding a bike. Google claims its text-to-image AI tool is more advanced than its competitor, DALL-E 2, but beneath the cute photos, there is an ugly truth about Imagen.

Since Imagen uses a pre-trained language model to generate realistic images, a potential risk is human bias, and the AI tool could mimic human behavior such as sexism and racism. Google acknowledges on its Imagen website that the AI tool “inherits the social biases and limitations of large language models," and therefore, the company will not release Imagen for public use.

Google also says that it doesn't portray people in the images because Imagen tends to reinforce gender stereotypes, evidenced by the fact that many AI creations Google has released only feature cute animals. Similarly, DALL-E-2 is restricted to a few users because of the same potential risks. Despite limitations, Google is working to improve its AI image tool in the future.