All AI-generated pics in Meta apps will soon be labeled with “AI info” text.
Meta will launch a technology on three of its social media arms. Once deployed, it will detect and label pictures by other companies’ AI tools.
Combating AI Fakery
If you don’t know, Meta already enforced AI detection on Facebook, Instagram, and Threads. It labels AI images generated by these social media platforms, but the ultimate goal is to create “momentum” to tackle AI fakery.
In a blog by senior executive Sir NIck Clegg, Meta will expand its labeling of AI fakes in the coming months. The company is currently working with industry partners on common technical standards for identifying AI content, audio, video, images, etc.
Full implementation will now show a text like “AI info” to avoid deceiving viewers. This is good because many are easily fooled by AI-generated images on social media. AI-generated images of heaven, hell, and afterlife have made their way into social media and tricked users that those were from scientific data or research.
Fake but not AI
Not everything fake is done by AI. According to the BBC, Meta’s Oversight Board criticized the company for its policy on manipulated media. This was a response to a ruling on a video of US President Joe Biden, which it was edited to make him look like he was touching his own granddaughter inappropriately.
Actual footage can be edited to represent or depict something differently. This is a workaround for anti-AI fakery and won’t easily violate related policies as long as the contents are real. Thus, policies must be updated regularly, whenever something arises to combat false information, fake news, and fraudulent content.
The upcoming AI labeling tech on Facebook, Instagram, and Threads may arrive as part of app and service updates.