If you search “Israel/Gaza” on Adobe stock images, you will be met with a flurry of pictures depicting the cities turning to rubble, cars enveloped in flames, and young children sitting and observing smoke rising from bombed areas. But what might not be immediately obvious about these images is that they are AI-generated. In December last year, Adobe began allowing users to put up AI-generated images for sale on its Adobe Stock platform. While Adobe does have a set of content guidelines under which users must ensure that they have the proper rights to list an AI-generated image for sale, the guidelines notably lack any mention of the ethics of creating AI-generated images on geopolitical conflicts.
How Adobe ensures transparency about AI-generated images:
A screenshot from Adobe Stock’s website
When you click on any of the images described above, you will see a small label that says that the image is AI-generated. Further, images generated using Adobe’s own AI image generator, Adobe Firefly, come with content credentials which is additional information added to an image during export or download. This information is stored in a dedicated, tamper-evident set of metadata. Even after such an image is screenshotted and manipulated further, the content credentials are still preserved. These credentials can be verified using Adobe’s Content Authenticity Initiative’s (CAI) website. CAI is an attempt to fight misinformation and “add a layer of verifiable trust to all types of digital content.”
Why it matters:
Despite Adobe’s efforts to combat misinformation, it seems to have missed out on the fact that the average person does not check metadata before they believe the authenticity of an image. This is especially significant as AI-generated images are already being used to spread propaganda about the conflict. These images make it harder for the rest of the world to truly know what is happening in the region, leaving them doubting real images as well.
Adobe’s response to these fake images:
In a comment to Vice, an Adobe spokesperson said that the images were, “labeled as generative AI when they were both submitted and made available for license in line with these requirements. We believe it’s important for customers to know what Adobe Stock images were created using generative AI tools.” The spokesperson further mentioned that the company is committed to fighting misinformation through CAI. Its statement notably lacks any mention of the potential impact these images could have on the conflict.
Note: The story was updated on November 8, 2023 at 3:47 to add more details to the headline.
Also read:
- Adobe Claims Firefly Application Can Avoid Copyright Infringement, Offers To Pay IP Indemnity To Businesses
- EU Opens Up An In-Depth Investigation Into Adobe’s Acquisition Of Interface Design App Figma
- Adobe Integrates Aadhaar KYC For Pushing E-Signatures; Where It Cannot Be Used
STAY ON TOP OF TECH NEWS: Our daily newsletter with the top story of the day from MediaNama, delivered to your inbox before 9 AM. Click here to sign up today!
