According to its quarterly threat report, Meta has taken action against propaganda campaigns on its platforms.
One such campaign used AI to manipulate political discourse and give the impression of greater support for specific viewpoints.
Meta Removes AI-Generated Disinformation Campaigns
Several campaigns promoted political narratives about ongoing events, with some originating from Israel and Iran expressing support for the Israeli government.
On Wednesday, May 29, Meta revealed that it had discovered content on its Facebook and Instagram platforms that was likely generated by AI.
This content was used deceptively, with comments appearing below posts from global news organizations and US lawmakers praising Israel's handling of the war in Gaza.
The social media company revealed that certain accounts on its platform had been disguised as various individuals, including Jewish students, African Americans, and other concerned citizens.
These accounts specifically targeted audiences in the United States and Canada. The campaign was credited to STOIC, a political marketing firm based in Tel Aviv.
Reuters said Meta has been uncovering basic profile photos created by artificial intelligence in influence operations since 2019.
However, this report is the first to reveal the utilization of advanced generative AI technologies since their emergence in late 2022.
Experts have raised concerns about the potential impact of generative AI, which can rapidly and inexpensively create text, images, and audio that closely resemble human creations.
There is apprehension that this technology could be exploited to spread disinformation and manipulate public opinion during elections.
Rampant Disinformation in Social Media Platforms During Elections
During a press call, Meta security executives expressed confidence in their ability to disrupt influence networks, stating that they did not believe novel AI technologies had hindered their efforts. Influence networks refer to coordinated attempts to push messages.
Executives stated that they had not come across AI-generated imagery of politicians that was so convincingly realistic that it could be mistaken for genuine photographs.
Many tech giants, including Meta, have been grappling with the challenge of addressing the potential misuse of new AI technologies, particularly in the context of elections.
It has come to light that image generators from major businesses like OpenAI and Microsoft have been found to produce photos containing voting-related disinformation despite these companies having strict policies against such content.
According to NDTV, many companies have strongly emphasized implementing digital labeling systems to mark AI-generated content from its inception.
However, these tools must be more capable of working with text, which has raised doubts among researchers regarding their overall effectiveness.
Meta will be tested as it faces upcoming elections in the European Union in early June and in the United States in November.
Join the Conversation