Meta is starting to make do with its promise to reduce disinformation on its platforms ahead of the 2024 Elections by expanding efforts to label AI-generated and AI-edited images more.

In a blog post on Tuesday, the social media giant announced that it has begun adding "photorealistic images" generated via Meta's AI feature with "visible markers."

An "invisible watermark" will also be inserted in its metadata to "help other platforms identify them."

For AI-generated images made outside of Meta's platforms, the company said it is currently working on a "range of options" to address the issue.

Among these options are ways to make it difficult to remove the markers and invisible watermarks, expand the application of its Community Standards, and require users to disclose if they used AI on the posted images.

Meta Collaborates with AI Research to Update Platform Policies

Meta said it will continue to work with its AI Research team to secure the technology from people aiming to use AI to further spread disinformation on its platforms.

The announcement comes in after Meta's oversight board called out the company for its "incoherent" rules against manipulated media.

Its oversight board earlier ruled not to remove the manipulated video of US President Joe Biden on Facebook on grounds that Meta does not have policies against non-AI-edited footage.

The platform has yet to issue a decision regarding the board's ruling, including a change in policies outside of AI-generated images.

Also Read: Meta Oversight Board Retains Manipulated Biden Video, Says Company's Rule 'Incoherent'

Election Disinformation on Meta Expected to Ramp Up

With almost half of the world entering the Elections season, Meta has also been ramping up its efforts to combat disinformation to prevent what happened during past voting periods.

Meta's platforms, particularly Facebook and Instagram, have been known to be vulnerable to disinformation during past elections.

At least 246.73 million Americans use Facebook in 2023, or 73.6% of the total population in the US. Many of these users are above the age of 18 and are qualified voters.

It did not help that Meta previously rescinded its decision to block out disinformation campaigns regarding the 2020 Elections, many of which denied the official result of the voting.

Related Article: Meta to Launch Fact-Checking on Threads Amid Rampant Misinformation