X Cracks Down on AI-Generated War Misinformation
X, formerly known as Twitter, is implementing stricter measures to combat the spread of deceptive artificial intelligence-generated videos, particularly those depicting the conflict-ridden Middle East. The social media giant will now suspend users from its monetization program for 90 days if they post AI-made war footage without clear disclosure. Repeat offenders face permanent removal from the program.
Nikita Bier, X’s head of product, announced the new policy, emphasizing the ease with which current AI technologies can create misleading content. “During times of war, it is critical that people have access to authentic information on the ground,” Bier stated. This move follows a surge in fabricated AI-generated posts on social media in the wake of recent geopolitical events in the region.
One particularly egregious example of misinformation involved a video purportedly showing Israeli soldiers in distress following an Iranian strike. This clip garnered over 1.4 million views. Another fabricated video, viewed by more than 2.1 million people, falsely depicted Dubai’s iconic Burj Khalifa engulfed in flames, claiming it was a result of an Iranian attack.
Similarly, a video circulating on X claimed to show “Iranian missiles hit[ting] central Israel,” featuring footage of a significant explosion. However, this clip was, in fact, AI-generated, though it was subsequently flagged by users on the platform. X has stated that AI-generated content will be identified through a combination of crowdsourced notes from users and metadata signals indicating the use of generative AI tools.
Another piece of misinformation falsely asserted that Iranian ballistic missiles had obliterated targets in Tel Aviv. The AI-generated footage showcased what appeared to be a barrage of rockets descending upon the Mediterranean city, complete with distant explosions and smoke clouds. In another instance, a video purported to capture an attack on an unnamed Israeli airport, presenting seemingly terrifying scenes that were entirely fabricated by AI.
Identifying AI-Generated Content
Distinguishing between authentic and AI-generated video content can be challenging, but several indicators can help users identify potential fakes. According to the BBC, common signs include low picture quality and unusually short video durations. Some AI bots may also incorporate outdated information, leading to inaccurate depictions of locations.
The Better Business Bureau points to other giveaways, such as strange textures or an overly “airbrushed” appearance. Physical inconsistencies, unnatural shadows, and unusual lighting can also be tell-tale signs of AI manipulation. Ironically, typos, while often indicative of human error, can sometimes be an encouraging sign in video content, as machines are less prone to making them.
Despite the challenges, Elon Musk has previously predicted that AI-generated video will dominate content consumption in the coming years. “Most of what people consume in five or six years – maybe sooner than that – will be just AI-generated content,” Musk remarked in October. This foresight highlights the growing importance of robust verification systems on platforms like X.
New Policy Implementation and Support
Under the updated guidelines, X users will be required to add an “Made with AI” label to their posts. This can be done by accessing the post’s menu and selecting “Add Content Disclosures.”
This initiative has received praise from various quarters. Sarah Rogers, the under secretary of state for public diplomacy, commended the move, stating, “This is a great complement to X’s community notes system, which results in less ‘reach’ (thus monetization) for content annotated as inaccurate.” She further added, “You don’t need a Ministry of Truth to incentivize truth online.”
The policy update reflects X’s ongoing efforts to strengthen its AI guardrails. Last month, the platform announced adjustments to its AI tool, Grok, to prevent the creation of overly sexualized images. Grok had previously faced criticism for generating content that touched upon antisemitic tropes and conspiracy theories.
The platform’s commitment to combating misinformation, especially in sensitive geopolitical contexts, underscores the evolving landscape of digital content creation and consumption. As AI technology advances, the need for transparency and accountability in online information becomes increasingly paramount.
