New Facebook Feature Uses AI to Edit Your Camera Roll Photos

11 hours ago 4

TEMPO.CO, Jakarta - Facebook is asking for permission to access users' phone camera roll to provide automatic suggestions in the form of AI-edited versions of photos, including images that have not been uploaded to the platform.

As quoted from Techcrunch on Friday, June 27, 2025, this feature appears when users are about to create a new story in the app. In the process, a notification will appear offering the option of "cloud processing" so that users can accept creative ideas from the system.

In the pop-up displayed, Facebook explains that by clicking "Allow," users grant access to create various creative ideas from their camera roll. These ideas can manifest as collages, automatic summaries, AI-assisted rearrangements, or specific photo themes. To enable this feature, Facebook will periodically upload content from the camera roll to its cloud server, using data such as capture time, location, or image theme.

The message emphasizes that only users can see the generated suggestions and that the media will not be used for advertising.

However, users who agree to this option are also agreeing to Meta's AI Service Terms. This gives permission for AI to analyze the content and facial features in photos, and to utilize information such as the date and presence of people or objects in the image to generate creative suggestions.

This feature shows how easily we can slip up when sharing personal data with AI technology providers. Like other major technology companies, Meta is competing in AI development, and the ability to utilize unpublished personal photos gives it an advantage. Unfortunately, users often consent to such features without fully understanding the consequences.

Based on Meta's AI Terms regarding image processing, users who share photos agree that Meta can analyze the images, including facial features, using artificial intelligence. This process allows Meta to provide new features, such as summarizing image content, editing, or generating new images based on the photo.

These terms also grant Meta the right to store and use users' personal information to adjust AI output. Meta states that human reviewers may examine user interactions with their AI system, including conversations.

However, they do not specifically explain the limitations of personal information, only mentioning that the data includes "requests, feedback, or other content" submitted by users.

It is still unclear whether photos uploaded in the "cloud processing" process are also included in the category of personal information that can be analyzed and stored. So far, there has been no strong public reaction to this feature. Some users began to realize the existence of this feature when creating Stories on Facebook, such as when Meta's AI automatically turned old photos into anime-style versions.

Some users in the anti-AI community on Facebook tried to find ways to turn off the feature. As it turns out, the settings can be found in the app's Preferences section, specifically in the "Camera Roll Sharing Suggestions" menu. There are two options: the first allows Facebook to suggest photos when browsing the app, and the second (cloud processing feature) allows Facebook to create AI-based images from the photos in the user's camera roll.

Meta's ability to access and use photos from the camera roll is not new. Some users have reported the existence of this feature since early this year, accompanied by screenshots of the pop-up that appears in Stories. Meta has also provided official guidelines for Android and iOS users to manage this feature.

Meta's AI Terms have been in place since June 23, 2024. However, it is difficult to track previous versions because there are no official public records, and the Internet Archive's archives are incomplete.

What is clear is that with this feature, Meta is going beyond just training AI based on publicly shared content. Now, privately sensitive data such as the contents of the camera roll are also being targeted. Users in the European Union were given until May 27, 2025, to express rejection of this policy.

Editor's Choice: Netflix Partners with NASA to Show Space Exploration

Click here to get the latest news updates from Tempo on Google News



Google Launches Doppl, an AI-Based Virtual Trial App for Clothing

1 hari lalu

Google Launches Doppl, an AI-Based Virtual Trial App for Clothing

Doppl, a new AI-based app from Google, is now available on iOS and Android devices


Larry Ellison Overtakes Jeff Bezos as the Second Richest Person in the World After Elon Musk

2 hari lalu

Larry Ellison Overtakes Jeff Bezos as the Second Richest Person in the World After Elon Musk

Founder of Oracle, Larry Ellison, overtakes Jeff Bezos as the second richest person in the world after Elon Musk.


Social Media Takes Over News Sites as Main Source of News as Trust Falls

3 hari lalu

Social Media Takes Over News Sites as Main Source of News as Trust Falls

A major global survey reveals that social media has overtaken news sites as main source of news amid falling trust in news


Four OpenAI Lead Researchers Leave for Meta, Slack Profile Deactivated

3 hari lalu

Four OpenAI Lead Researchers Leave for Meta, Slack Profile Deactivated

Profile of the four researchers from OpenAI has been deactivated on Slack after they left the company to join Meta.


Telegram Founder Pavel Durov Plans to Leave Fortune to Over 100 Children He Fathered

6 hari lalu

Telegram Founder Pavel Durov Plans to Leave Fortune to Over 100 Children He Fathered

The founder of the instant messaging app Telegram, Pavel Durov, revealed that over 100 of his biological children will inherit his wealth.


LinkedIn CEO Admits AI Writing Feature Struggles to Attract Users

6 hari lalu

LinkedIn CEO Admits AI Writing Feature Struggles to Attract Users

LinkedIn CEO stated that the AI-powered writing tool intended for creating resumes and posts on the platform has not been widely adopted by users.


MIT Researchers: ChatGPT Use May Lead to Cognitive Decline

7 hari lalu

MIT Researchers: ChatGPT Use May Lead to Cognitive Decline

A study reveals that the use of the LLM models like ChatGPT can erode critical thinking skills, creativity, and decrease memory capabilities.


U.S. House Bans WhatsApp from All Staffers' Devices

7 hari lalu

U.S. House Bans WhatsApp from All Staffers' Devices

The U.S. Cyber Security Office has categorized WhatsApp as a high-risk application.


Apple Reportedly in Talks to Acquire AI Startup Perplexity

8 hari lalu

Apple Reportedly in Talks to Acquire AI Startup Perplexity

Apple's rumored acquisition of Perplexity is said to be the largest in the company's history, surpassing the $3 billion purchase of Beats in 2014.


Overcast Weather Expected in Greater Jakarta Today, BMKG Says

8 hari lalu

Overcast Weather Expected in Greater Jakarta Today, BMKG Says

BMKG reports that significant cloud growth is possible across Indonesia over the next week.


Read Entire Article
Parenting |