もっと詳しく

Adobe is the latest stock image provider to take a stance on AI-generated art. The company has updated its guidelines to allow generative AI artwork on its stock photo service as long as it meets certain criteria. Contributors will have to label any AI-made content, and they’ll require permission for any reference images or text prompt used to produce the art. Creators will likewise need releases for any material depicting recognizable people.

The new policy also warns producers against misusing AI by submitting multiple images based on the same prompt. They can’t use misleading, repetitive or vague descriptions, and must submit their works as illustrations (not pictures) even if they’re photorealistic. As with regular images, Adobe promises indemnification in the event there’s an intellectual property dispute. You shouldn’t be in deep trouble if you unwittingly use items that infringe someone else’s rights.

Adobe is effectively trying to strike a balance between embracing new technology and avoiding copyright issues. Other stock photo providers have taken wildly varying approaches. Getty Images banned AI-generated imagery over rights concerns, while Shutterstock has teamed with DALL-E creator OpenAI to sell algorithm-based images. In some cases, tool developers have avoided wading into the debate — Google won’t offer Imagen to the public until it believes there’s a “responsible” way to do so. 

This move might not lead to a flood of AI-made images to use for your next presentation or website. However, it could be helpful if you’re eager to use unconventional content without worrying about the risks of unexpected lawsuits or royalty payments. If nothing else, Adobe’s move could increase acceptance of AI stock art by making it available to a wider audience.