By now you’ve probably tried one of the new AI-based image generation tools. This tool “samples” various image repository websites and online references and creates all new visuals based on your text prompts.

Dar E is the best known of these new apps, but Midjourney has also gained popularity in recent months, allowing users to create stunning visual artwork with virtually no effort.

But what are the usage rights for the visuals you create, and can marketers actually use these images in their content? No potential copyright issues.

There are some conditions to consider, but at the moment it seems possible.

According to the terms of use Dar Eusers have the right to use their work for any purpose, including commercial use.

“Subject to your compliance with these terms, content policy, you may use Generations for any lawful purpose, including commercial use. This means that you can sell the rights to the generation that created it and incorporate it into your work, such as books, websites, presentations, or even commercialize it.”

Yes, you can sell the visuals you create, but most stock photo platforms currently Re-evaluate whether you would actually accept such a sale.

this week, Getty Images has become the latest platform to ban uploading and selling illustrations generated by AI art tools. Here’s why, according to Getty.

“…the copyright concerns of the output from these models and the unresolved rights issues regarding the images, the image metadata, and the individuals contained in the images”

One concern here is that the visuals used as source material for these AI-generated depictions may not be licensed for commercial use.

Still, it’s not necessarily a decisive legal barrier.

As explained by The Verge:

“Software like Stable Diffusion [another AI art tool] I am trained on copyrighted images I collect from around the web, including personal art blogs, news sites, and stock photo sites like Getty Images. Scraping is legal in the US, and software output seems to be covered by the “fair use” doctrine. However, fair use weakens protection against commercial activities such as selling photos. Also, some artists whose work has been scraped and imitated by companies that create AI image generators, call for new laws Regulate this domain.”

In fact, various proposals have been put forward to potentially regulate and even limit the use of these tools to protect artists. However, such rules are not yet in place and it could be years before a legal consensus is established on how to better protect the artists whose work is sourced on the backend. there is.

There are even questions about the technical process of creation, and how it applies to legal protection in this sense. United States Copyright Office It effectively implies that AI-generated images cannot be copyrighted, as they require an element of “human copyright”.

Regarding specific content policies, DALL・E’s terms of use state: Use the app to “create, upload, or share images that are not G-rated or potentially harmful.”

So while there are no depictions of violence or hate symbols, the DALL-E team encourages users to: Proactively disclose AI involvement in content.

Additional guidelines for DALL E are:

• Do not upload images of people without their consent.
• Do not upload images for which you do not have proper usage rights.
• Do not create images of public figures.

A more complicated issue can arise here. As JumpStory pointed out, users of AI image generation tools should be aware of potential copyright issues when attempting to create images that include real people.

JumpStory states that many of the source images for the DALL-E project are actually from Flickr and are subject to Flickr’s terms of use. This is fine for most generated depictions such as landscapes and artwork, but one of these tools may end up using a person’s true face. Misrepresentation, depending on the context.

Again, the legal details here are complicated, and in practice it is unclear how such cases would actually be prosecuted, as there is no true precedent to continue. If you are trying to generate an image of a person, it can get complicated if the visual directly resembles a real person.

Clearly stating that the image was generated by AI almost always provides some clarity. However, as a precaution, it may be safer to avoid people’s faces clearly appearing in the resulting image.

MidJourney’s terms also make clear that infringement of intellectual property rights will not be tolerated.

“If you knowingly infringe on someone else’s intellectual property and it costs us money, we will find you and collect the money from you. Do not do that.

It’s a strangely grim story about legal documents, but the impetus is clear. While you can use these tools to create art, creating images that are clearly derivative or infringing intellectual property rights can be problematic. is recommended.

But in practice, from a legal point of view, this is the situation. These systems take elements from other visuals online, but the actual images they create didn’t exist until they created them, so they aren’t subject to copyright. , effectively the original source.

At some stage, such legal technology may change. Also, I’m guessing someone will host an AI art show of his or something, or sell a collection of his AI-generated art online depicting key elements of other artists’ work. increase. , and it would spark new legal debates about what constitutes intellectual property infringement in this regard.

However, at this time, it is mostly fine to fully use the images produced by these tools, subject to the terms stated in the documentation of the tools themselves.

Note: This is not legal advice. We encourage you to check with your legal team to clarify such company’s stance before proceeding.