At the beginning of the year, many people were worried about the potential dangers of generative AI when it came to global elections. They feared that this technology could be used to mess with elections by spreading false information and propaganda. However, as the year comes to a close, the company Meta—which owns platforms like Facebook, Instagram, and Threads—says that those fears didn’t really come true, at least not on their platforms. In a recent blog post, they explained that the use of generative AI had very little influence during major elections.
Meta reviewed content related to important elections in various countries, including the United States, Bangladesh, Indonesia, India, Pakistan, the European Union, France, the United Kingdom, South Africa, Mexico, and Brazil. They found that, while there were some cases where AI was confirmed or suspected to be used for spreading misinformation, the overall amount was quite low. They stated that their existing rules and protective measures were effective in keeping the risks associated with generative AI content under control.
Specifically, during the election periods in those countries, AI-generated content that was flagged or fact-checked as misinformation accounted for less than 1% of all the misinformation they reviewed. For example, Meta’s Imagine AI tool, which generates images, blocked around 590,000 requests to create fake images of well-known political figures during the month leading up to election day. This was an effort to prevent the spread of deepfake images that could deceive voters.
Meta also found that groups of accounts that aimed to spread misinformation didn’t use generative AI very successfully. They reported only small advancements in creating content using AI. The company explained that their approach focuses on the behavior of these accounts rather than the specific content they share, whether it’s created by AI or not. This focus helped them continue to shut down these misleading campaigns effectively.
Read Also: The Race Is On To Have AI Agents Shop Online For You
Over the year, Meta reported that they took down around 20 covert influence operations that were trying to manipulate public opinion and interfere from abroad. Most of these networks didn’t have genuine followers and relied on fake likes and followers to give the impression of being more influential than they actually were.
Meta also pointed out that while they were combating misinformation, other platforms like X (formerly Twitter) and Telegram were often where false videos about the U.S. election—linked to Russian influence operations—were being shared.
As they reflect on everything that has happened over the year, Meta assured users that they will continue to review their policies and make any necessary changes in the coming months.
