Clarification on generative AI submission guidelines

[Mat Hayward]

As generative AI evolves, we want to share a clarification to our generative AI submission guidelines. We prohibit generative AI content with titles that imply that the content depicts an actual newsworthy event from being accepted to Adobe Stock. For more information, please see our recent blog post and updated guidelines.

https://blog.adobe.com/en/fpost/adobe-stock-updates-to-combat-deceptive-content-misuse

https://helpx.adobe.com/stock/contributor/help/generative-ai-content.html

Let me know if you have any questions,

Mat Hayward

[Jill_C]

Thanks for the clarification. This has been a frequent subject of discussion in the forum.

[Nancy OShea]

We prohibit generative AI content with titles that imply that the content depicts an actual newsworthy event...

By @Mat Hayward

========

It's a shame this has to be spelled out for people.  

[Jill_C]

Indeed. I have actually had online arguments with AI contributors who see no harm in "pretending" that their creations can represent real world, actual events.

[Nancy OShea]

Indeed. I have actually had online arguments with AI contributors who see no harm in "pretending" that their creations can represent real world, actual events.

By @Jill_C

==========

Or people, places & landmarks.

AI is fake.  It will always be fake.  There is nothing worse than labeling fake images as real.  At best, it's misleading & deceptive.  At worst, it's intentionally dishonest. 

[Abambo]

@Mat Hayward @MatHayward ,

could you please pin this, as it should be visible fore some time. 

[MatHayward]

Done, thanks Abambo.

-Mat 

[daniellei4510]

It's also a good idea not to include "A photograph of..." in the title or "photography" or "photo" in the keywords, even if it looks like a real photo and not an illustration. At least that's a practice I try to maintain.

[Kini-II]

For me, this means that I'm not uploading any more AI images at the moment, as I'm waiting to see what form Adobe wants the information for their content credentials to take. As JPG metadata or something else. Above all, I then don't have many images in the approval queue, to which I don't have editable access. How it works with the already accepted AI images is unfortunately not clear from the information, at least I can't find any information about it.

[Abambo]

You do not understand the message here. You simply can't pretent your generative AI would be in relation with a certain real event event. Like producing generative AI and calling it after a real event. Like a war in some location like Ukraine or Palestine. 

[Kini-II]

i understand everything and i never did something wrong. Never i described a AI as it would be real. But i do not understand how this content credential thing works technically. I am only afraid to do something wrong. If adobe tells me to add my AI Tool to the jpg metadata i will do it. If they tell ne i have to describe it manually i will do it. But at the moment i do not know what to do. Belongs it to a ai picture of a tea cup also? I want to wait that's all. 

[Abambo]

You do as you please. Generative AI images are getting earmarked as generative AI. But if you submit your assets as a newsworthy asset as if it would depict a real picture (even if marked as generative AI), it would be refused. If you do that massively, I suppose, you will earn an account suspension. 

[Giovanni33881797518g]

Sounds fair and good to me. I appreciate your responsible and reasonable approach to AI that is not penalizing AI artists who play a fair game and work hard to produce great images for your clients.

[Bhakiag340862590x9u]

Certainly, when submitting content or using generative AI, it's important to adhere to ethical guidelines and consider the potential impact of generated material. Here are some general guidelines:

Respect Privacy: Avoid using or generating content that includes personally identifiable information without explicit consent.

Avoid Harmful Content: Refrain from creating or sharing content that promotes harm, violence, discrimination, or illegal activities.

Follow Copyright Rules: Ensure that generated content doesn't infringe on copyrights or intellectual property rights. If the input data includes copyrighted material, make sure your use complies with relevant laws.

Maintain Ethical Standards: Consider the ethical implications of the content being generated. Be mindful of the potential consequences and impact on individuals or communities.

Consent for Personal Data: If using personal data for training or generating content, make sure you have proper consent and adhere to data protection regulations.

Transparency: If the generated content is used in public-facing applications, consider providing transparency about its AI-generated nature to users.

These guidelines are meant to promote responsible and ethical use of generative AI. Always be aware of the ethical implications of your actions and strive to use the technology in a manner that respects individuals and society as a whole. Additionally, it's recommended to be familiar with the specific terms of use and policies of the platform or service you are using for generative AI.

[Marsh292012685ro1]

So, suppose if I'm a news channel and I try to be innovative bu bringing adobe generative AI images and I want to use adobe images but in your article:https://blog.adobe.com/en/fpost/adobe-stock-updates-to-combat-deceptive-content-misuse

You have mentioned such words "But I want to be clear: Adobe Stock content is not real-world journalistic content and should never be used in journalistic mediums to depict newsworthy events."

So, it means that it'll not be appropriate to use generative images for my channel as it'll be misleading?

[Abambo]

You should make clear that the asset, your showing is not a real asset, if it could be taken for a real one. Deceptive pictures are not new, btw. It just got easier now to create them.

[Irakli Topuria]

Hi Abambo. It was declared a while ago, that there is no need to mention "Generative AI" in titles or keywords. So is it back? Should we mention "Generative Ai" in titles and keywords again?

Thank you

[Abambo]

No, you have a flag to set during the submission.

[Irakli Topuria]

Sure, I know that, but as you mentioned down below, Adobe is now accepting generative Ai in vector graphics, so are these checkmarks also available when you upload Ai vectors?

Thanx

[Jill_C]

Yes, you have to select the check box that indicates "Created using Generative AI tools".

[Irakli Topuria]

Thank you!

[Abambo]

Sure, I know that, but as you mentioned down below, Adobe is now accepting generative Ai in vector graphics, so are these checkmarks also available when you upload Ai vectors?

Thanx

By @Irakli Topuria

As Adobe is accepting generative AI as vectors, the checkmarks are also available with vector files. You can test for yourself. Upload any vector file, check the buttons, delete that vector. You do not need to submit to confirm the presence.

[Irakli Topuria]

Thanks a lot!

[Abambo]

You're welcome.

[Jill_C]

AP (Associated Press), the definitive source of guidelines for journalistic integrity, has already issued guidance in the use of AI in journalism:

https://blog.ap.org/standards-around-generative-ai#:~:text=However%2C%20if%20an%20AI%2Dgenerated,as%....

• In accordance with our standards, we do not alter any elements of our photos, video or audio. Therefore, we do not allow the use of generative AI to add or subtract any elements.

No one who calls themselves a "journalist" would use AI images to accompany a news article. That's also why it's important for images to have content credentials to verify that Gen AI tools were not utilized to create or alter the image.