Adobe Adds Generative AI Features to Photoshop Software 

Figurines are seen in front of the Adobe logo in this illustration taken June 13, 2022. (Reuters)
Figurines are seen in front of the Adobe logo in this illustration taken June 13, 2022. (Reuters)
TT

Adobe Adds Generative AI Features to Photoshop Software 

Figurines are seen in front of the Adobe logo in this illustration taken June 13, 2022. (Reuters)
Figurines are seen in front of the Adobe logo in this illustration taken June 13, 2022. (Reuters)

Adobe Inc on Tuesday said it is adding artificial intelligence (AI) technology for generating imagery into Photoshop, its flagship software for editing images.

The San Jose, California-based company said it was the start of a major push to add such AI technologies to its suite of programs aimed at creative professionals.

While programs such as OpenAI's Dall-E have captured the public imagination by transforming text prompts into images, they have not yet seen wide use by big corporations because of legal questions around the data used to develop the systems.

Adobe has sought to address those concerns with a core technology system it calls Firefly, which was specifically created with legal-to-use image data and which Adobe says can be used in commercial settings.

Adobe has been testing the system for about six weeks on a standalone website and on Tuesday said it will add features based on it to Photoshop, perhaps the company's best known product.

One new feature will be called "Generative Fill" and it will allow users to extend an original image that was cropped in too closely with computer-generated content, or add features based on a text description.

The feature can, for example, take a picture of a single flower and turn it into a field of flowers with a mountain range behind it.

Ely Greenfield, chief technology officer for digital media at Adobe, said that the intention of the tool is not to replace graphic artists but to make it faster for them to create new images out of multiple ideas. In the past, they would have had to spend valuable hours searching photo archives and stitching together pieces of existing images by hand.

"This just dramatically accelerates that production work," Greenfield said.



Meta to Start Using Public Posts on Facebook, Instagram in UK to Train AI

Meta AI logo is seen in this illustration taken May 20, 2024. (Reuters)
Meta AI logo is seen in this illustration taken May 20, 2024. (Reuters)
TT

Meta to Start Using Public Posts on Facebook, Instagram in UK to Train AI

Meta AI logo is seen in this illustration taken May 20, 2024. (Reuters)
Meta AI logo is seen in this illustration taken May 20, 2024. (Reuters)

Meta Platforms will begin training its AI models using public content shared by adults on Facebook and Instagram in the UK over the coming months, the company said, after it had paused the training in the region following a regulatory backlash.

The company will use public posts including photos, captions and comments to train its generative artificial intelligence models, it said on Friday, adding that the training content will not include private messages or information from accounts of users under the age of 18.

The update follows Meta's decision in mid-June to pause the launch of its AI models in Europe after the Irish privacy regulator told the company to delay its plan to harness data from social media posts.

The company had then said the delay would also allow it to address requests from Britain's Information Commissioner's Office (ICO).

"Since we paused training our generative AI models in the UK to address regulatory feedback, we've engaged positively with the ICO ... this clarity and certainty will help us bring AI at Meta products to the UK much sooner," Meta said on Friday.

Facebook and Instagram users in the UK will start receiving in-app notifications from next week explaining the company's procedure and how users can object to their data being used for the training, Meta added.

In June, the company's plans faced backlash from advocacy group NOYB, which urged national privacy watchdogs across Europe to stop such use of social media content, saying the notifications were insufficient to meet EU's stringent EU privacy and transparency rules.