Meta’s Oversight Board is calling on the firm to do more to prevent the spread of misinformation across its platforms after an AI-generated video claimed to show footage of the current conflict between the US, Israel, and Iran.

The Oversight Board specifically chastised the firm for leaving up a video that claimed to show damage enacted by Iranian forces on an Israeli city, which was actually generated by AI.

The ongoing conflict in the Middle East has seen rampant misinformation and disinformation proliferate online, and social media profiles share either AI-generated content or content from other conflicts claiming to show current events.

The board is calling on Meta to overhaul its AI rules to halt the “proliferation” of fake AI-generated content which “challenged the public’s ability to distinguish fabrication from fact” which is says is “risking a general distrust of all information.”

In response, Meta said that it would label the video as AI within seven days.

The Board, which is semi-independent from Meta and tasked with supervising content moderation across Facebook, Instagram, and WhatsApp.

The recent issue concerning the AI-generated video is a mere reflection of “inefficiencies in Meta’s current approach during armed conflicts,” the board said.

It has said that Meta should “much more frequently” label AI content proactively, instead of relying on “self-disclosure” or user complaints, which is its current approach.

It called Meta’s current approach “neither robust nor comprehensive enough to contend with the scale and velocity of AI-generated content, particularly during a crisis or conflict where there is heightened engagement on the platform.”

Previous issues of AI-generated content concerning conflict and war had raised alarm bells to the Oversight Board, which found that Meta had neglected to act on user complaints until the issue was raised specifically to the Board.

The Oversight Board found that Meta’s defence, saying that content did not “directly contribute to the risk of imminent physical harm” to be an unrealistic and overly high bar for content removal.

Recommended reading

• AI-driven Misinformation Tops WEF Short-term Threat List
• Meta To Start Labelling AI-generated Images
• EU Calls Out Meta & TikTok Amid Flurry of Disinformation Claims
• EU Investigating X Over Alleged Disinformation Amid Israel-Hamas Conflict

Meta has responded to the Board’s insistence it must “do more to address the proliferation of deceptive AI-generated content on its platforms”, saying that it will follow the board’s suggestions in a statement.

However, Meta said that it would follow the Board on cases with “identical content” concerning the “same context” as the video the board had analysed.

Meta has faced allegations of spreading dangerous content during conflicts even before the proliferation of AI videos. The Rohingya, an ethno-religious minority in Myanmar, sued Facebook in 2021 alleging the company failed to prevent posts that incited violence against them during the genocide in the group faced in the country.

A BBC Wales investigation recently revealed deepfake content farms were spreading misinformation surrounding UK politics proliferating across social media sites.