Meta has announced that it will launch an Elections Operations Center in the European Union (EU) ahead of the elections for the European Parliament in June to identify potential threats and put mitigations in place in real time. The company says that it works with 26 fact-checking organizations in the EU who review and rate content on Meta’s platforms and that artificial intelligence (AI) generated content would also be reviewed by said fact-checkers.
One of the rating options that will be made available to the fact-checkers is ‘altered’, which includes, “faked, manipulated or transformed audio, video, or photos.” When a piece of content is rated as altered, Meta adds a fact-checked label on it and then reduces its distribution on people’s feeds, making it less likely to be seen. The company mentioned that it labels realistic-looking content generated by MetaAI and also plans to introduce tools to label AI-generated images from Google, OpenAI, Microsoft, Adobe, Midjourney, and Shutterstock that users post to Facebook, Instagram, and Threads. It will also add a feature for people to disclose when they share AI-generated video or audio so that it can label it as such. Those that fail to disclose AI-generated content may be penalized.
Advertisers who run ads related to social issues, elections, or politics on Meta’s platforms are required to disclose if they use a photorealistic image or video, or realistic-sounding audio, that has been created or altered digitally, including with AI, in certain cases.
This announcement comes two weeks after a group of 20 AI companies including Adobe, Microsoft, Meta, Google, and TikTok signed an accord to combat the deceptive use of AI in the 2024 elections.
Will Meta’s steps to curb AI-generated misinformation help?
Deep fake detection has emerged as a grave concern across the world ahead of the upcoming elections in various jurisdictions. While Meta has talked about introducing tools to label AI content, there is a possibility that this might not be the most effective step. The company has not disclosed how it would label content but given the sheer scale of its platforms it would need to use AI detection algorithms in sweeping posts and adding the labels. And these detectors can go wrong.
Speaking at MediaNama’s discussion on deep fakes and democracy researcher from UC Berkeley Gautham Koorma pointed out that AI detectors can end up making mistakes that have grave consequences. For example, one of the popular AI detector tools got images from the war in Gaza wrongly labeled as a deep fake, and multiple prominent news outlets carried this false flag in their reporting. Even in cases where the algorithm is 90 percent accurate, it is not enough because you are getting a false detection in 10 out of 100 cases. This false detection rate becomes extremely high when considering the billions of posts that get put up on social media platforms daily. “So, when you’re using these algorithms on social media for automated detection of deepfakes you want to get 99.999% accuracy,” Koorma had explained. Creating a detection tool with such a high accuracy would pose its own challenges for the company.
Other key steps taken by the company :
- Meta says that it will make it easier for fact-checkers to debunk information ahead of the EU elections. It plans to use keyword detection to group pieces of content together and also intends to onboard its research tool called Meta Content Library.
- Ads posted on Meta’s platforms would not be allowed to use debunked content.
- The company is working with the European Fact-Checking Standards Network (EFCSN) on a project to help train fact-checkers across Europe on the best way to evaluate AI-generated and digitally altered media, and on a media literacy campaign to raise public awareness of how to spot that type of content.
Also read:
- Tech Companies Sign An Accord To Curb Deepfakes Ahead Of 2024 Elections
- Explainer: Why Detecting Deepfakes Is A Challenging Problem #NAMA
- TikTok To Detect And Label AI-Generated Content Ahead Of US 2024 Elections
STAY ON TOP OF TECH NEWS: Our daily newsletter with the top story of the day from MediaNama, delivered to your inbox before 9 AM. Click here to sign up today!