Meta Says AI-Generated Misinformation Has Been Low for 2024 Elections

Thursday - May 23 2024, 06:17 UTC - 6 months ago

Meta states that there has been minimal AI-generated misinformation in the 2024 elections. Experts are concerned about AI-generated disinformation interfering with elections, and Meta has faced criticism for its content moderation policies in the past. The company is now utilizing fact checkers and AI technology to combat election interference, although these tools are not yet perfect. Clegg defends Meta's decision to allow ads claiming the 2020 US election was stolen and states that AI is an important tool in identifying and removing false information.

Meta, formerly known as Facebook, has stated that it has seen strikingly little AI-generated misinformation around the 2024 elections despite major votes scheduled in countries such as Indonesia, Taiwan, and Bangladesh. Nick Clegg, the company's president of global affairs, emphasized that while AI-generated interference is present, it is not happening at a significant volume or on a systemic level. Clegg mentioned that Meta has seen attempts at interference in the Taiwanese election, but the scale of the attempted interference is manageable.

As more than 50 countries prepare for elections this year, experts have expressed concerns about the potential for AI-generated political disinformation to interfere with the democratic process. The fear is that malicious actors will utilize generative AI and social media to spread false information and manipulate voters.

Meta, a company that has previously faced criticism for its content moderation policies during elections, is now ramping up its efforts to combat election interference. Clegg stated that the company has removed over 200 networks of coordinated inauthentic behavior since the 2016 US presidential election. In order to identify unwanted groups on its platforms, Meta now relies on a combination of fact checkers and AI technology. However, Clegg admits that these tools are still imperfect and immature, particularly when it comes to detecting AI-generated content. Watermarks, which are used to verify the authenticity of media, are not widely adopted in the AI industry and are easy to tamper with. This poses a challenge in identifying and removing AI-generated text, audio, and video that contain misinformation and disinformation.

Despite these challenges, Clegg reassures that AI is an important tool in the fight against election interference. He also defends Meta's decision to allow ads claiming that the 2020 US election was stolen, stating that similar claims are common in elections around the world. He adds that it is not feasible for Meta to relitigate past elections and that the company's systems should be able to detect and remove false information regardless of its source. However, this decision has been met with criticism, with eight state secretaries of state writing a letter to Meta CEO Mark Zuckerberg, expressing concerns about the potential danger of these ads in further threatening public trust in elections and the safety of election workers.