Meta’s AI-generated misinformation efforts for the 2024 elections
Meta has seen strikingly little AI-generated misinformation around the 2024 elections despite major votes in countries such as Indonesia, Taiwan, and Bangladesh, said the company’s president of global affairs, Nick Clegg, on Wednesday.
“The interesting thing so far—I stress, so far—is not how much but how little AI-generated content [there is],” said Clegg
during an interview at MIT Technology Review’s EmTech Digital conference in Cambridge, Massachusetts.
“It is there; it is discernible. It’s really not happening on … a volume or a systemic level,” he said. Clegg said Meta has seen attempts at interference in, for example, the Taiwanese election, but that the scale of that interference is at a “manageable amount.”
As voters will head to polls this year in more than 50 countries, experts have raised the alarm over AI-generated political disinformation and the prospect that malicious actors will use generative AI and social media to interfere with elections. Meta has previously faced criticism over its content moderation policies around past elections—for example, when it failed to prevent the January 6 rioters from organizing on its platforms.
Clegg defended the company’s efforts at preventing violent groups from organizing, but he also stressed the difficulty of keeping up. “This is a highly adversarial space. You play Whack-a-Mole, candidly. You remove one group, they rename themselves, rebrand themselves, and so on,” he said.
Clegg argued that compared with 2016, the company is now “utterly different” when it comes to moderating election content. Since then, it has removed over 200 “networks of coordinated inauthentic behavior,” he said. The company now relies on fact checkers and AI technology to identify unwanted groups on its platforms.
Earlier this year, Meta announced it would label AI-generated images on Facebook, Instagram, and Threads. Meta has started adding visible markers to such images, as well as invisible watermarks and metadata in the image file. The watermarks will be added to images created using Meta’s generative AI systems or ones that carry invisible industry-standard markers. The company says its measures are in line with best practices laid out by the Partnership on AI, an AI research nonprofit.
But at the same time, Clegg admitted that tools to detect AI-generated content are still imperfect and immature. Watermarks in AI systems are not adopted industry-wide, and they are easy to tamper with. They are also hard to implement robustly in AI-generated text, audio, and video.
Ultimately that should not matter, Clegg said, because Meta’s systems should be able to catch and detect mis- and disinformation regardless of its origins.
“AI is a sword and a shield in this,” he said.
Source: www.technologyreview.com