Adobe Photoshop will now give users access to generative artificial intelligence (AI) tools that will enhance their ability to alter images.
Key Details
- The ability to make fake images with AI tools has been a concern since the emerging technology first became available.
- Deep fakes and highly-realistic fake images have the potential to circulate misinformation and create pictures of public figures in compromising positions.
- While photoshopped images today are fairly easy to detect with the human eye, critics worry that AI technology is developing faster than the human ability to spot it.
Why it’s news
Already fake images have caused a stir. Doctored photos of the Pentagon ablaze after an explosion were shared on Twitter. Before the photos were revealed as fakes, the markets reacted with a brief dip, Axios reports. Without the ability to determine the truth of these images, AI-generated pictures have the potential to do a lot of harm.
Because of these concerns, Adobe’s announcement has been met with criticism. While the company is not directly promoting false images, it gives its users the tools to create hyper-realistic fakes.
Photoshop’s “generative fill” tool will allow users to more easily add or remove objects and change backgrounds. This tool has many available uses, including time-saving for professional photographers. Users can tell the AI tool to replace the existing background with a preferred option or remove or eliminate objects, such as removing people from the background of an image.
AI can also fill in blank spaces to “uncrop” a photo. If the subject is too close to the camera, AI can guess what the rest of the background would look like and extend the image. These features can be extremely useful for photographers who can change their customers’ settings from a grassy field to the top of a mountain, all with a few clicks of the mouse.
The AI tool is still in the beta version of Photoshop, but the company plans to include it with the official version later in the year. It will also be free on Adobe’s Firefly website, Axios reports.
In response to critics, Adobe says that those who plan to use this technology for deep fakes already have access to the tools. Unlike other generative AI image-creators, Adobe’s tool will not create images of public figures and celebrities.
While Adobe has many critics, its chief technology officer Ely Greenfield is right, “We’re already in that world.” Those who want to make deep-fake images already have the knowledge and the tools to do so. The public will need a better understanding of deep fakes and how to spot them.
Adobe says the public needs a better understanding of what type of image generation is possible. To support this, Adobe is promoting its Content Authenticity Initiative, a coalition announced four years ago. With this tool, users can track how and when an image changes.