After a series of artificial intelligence text-to-image generators that came into the picture one by one, the Australian design application Canva is the latest one to launch its own AI text-to-image generator. The company began testing the feature in September and has now rolled it out to around 100 million users on the application.
With this tool, Canva has enabled the generation of 100 images in a day on its free as well as paid versions of the application. The Co-Founder and CPO of Canva, Cameron Adams, wrote in a blog post announcing the release, “Our mission has always been to make complex things simple, and tapping into exciting new ways to empower our community to bring their ideas and imagination to life is an important part of that.”
Our mission has always been to make complex things simple, and tapping into exciting new ways to empower our community to bring their ideas and imagination to life is an important part of that.
Stable Diffusion open-source has been integrated to enable this feature on Canva. When you check out the text-to-image feature on Canva, you will find a text box where you can input your text prompt and a couple of suggestions listed to get you started with the generations. After you have generated your desired image, you also have the option to use numerous photo styles like painting, photo, drawing, pattern, etc. to edit the generated image.
With a new text-to-image tool making it to the news every day, it circles us back to whether appropriate filters are in place to ensure the tool’s safe. The company also said in the blog, “We’ve invested heavily in safety measures that help the millions of people using our platform ‘be a good human’ and minimize the risk of our platform being used to produce unsafe content.” The tool also has a report option that users can use to ensure that the generations produced do not cause any kind of harm.
Recently, there has also been noise around the use of copyrighted images in the training data of text-to-image tools. In an interview with The Verge, the Co-Founder said, “I think there are legitimate questions about the extent to which AI productions can be considered fair use, and that’s going to differ around the world. We’re keeping a close eye on it, but it’s all still pretty up in the air.”
We give ownership of the images to the users, but we don’t claim that they can be copyrighted by those users.
He also added that the company is working closely with its contributors and users to try and answer questions related to copyright. Speaking about the generated images, he added, “We give ownership of the images to the users, but we don’t claim that they can be copyrighted by those users.”
The company has clarified that these images can be used for personal as well as commercial purposes until users follow the company’s terms. They also mentioned that although generations can be used, the user doesn’t have exclusive rights to it and it can be used by other people.