Meta to begin labelling AI-generated content from May to counter manipulated media

Meta will begin labeling a wider range of video, audio and image content as "Made with AI" on its platforms, including Facebook and Instagram starting in May to counter manipulated media, according to its official statement.

Meta's new policy is a response to feedback from its independent Oversight Board, which urged an update to the current policy.

"We are making changes to the way we handle manipulated media based on feedback from the Oversight Board and our policy review process with public opinion surveys and expert consultations," Meta said.

The labels could be generated through self-disclosure when a user posts content, as a result of advice from fact-checkers, or via Meta detecting invisible markers of AI-made content.

"We will begin labelling a wider range of video, audio and image content as "Made with AI" when we detect industry standard AI image indicators or when people disclose that they're uploading AI-generated content," it added.

Meta also underscored the usage of realistic AI-generated tools to create content like audio and photos and its evolution at present.

"In the last four years, and particularly in the last year, people have developed other kinds of realistic AI-generated content like audio and photos, and this technology is quickly evolving. As the Board noted, it's equally important to address manipulation that shows a person doing something they didn't do," it said.

New AI video tools increase worries of deepfakes ahead of elections. Experts worry malicious actors could use AI tools to create deepfakes, confusing and misleading voters in an election year. The video that OpenAI released to unveil its new text-to-video tool, Sora, has to be seen to be believed.

Earlier last month, in a candid conversation with Microsoft co-founder, Bill Gates, Prime Minister Narendra Modi addressed the challenges posed by AI and underlined the significance of initially using watermarks on AI-generated content to make users aware and prevent misinformation.

"Addressing the challenges AI presents, I have observed that without proper training, there's a significant risk of misuse when such powerful technology is placed in unskilled hands. I've engaged with leading minds of AI, I suggested that we should start with clear watermarks on AI-generated content to prevent misinformation.

This isn't to devalue AI creations but to recognise them for what they are," PM said.

"Data security remains a paramount concern and while India has a legal framework in place,. For instance, I have initiated storing all university certificates in the cloud to reduce costs and fulfil various needs efficiently. Previously, there was a push for stringent compliance requirements. I advocated for simplification, suggesting that data be uploaded to the cloud with shared IDs, enabling us to access necessary information directly. This approach underscores my commitment to leveraging technology not just to enhance services but to significantly improve the ease of living for our citizens," he added.

Bill Gates also acknowledged the challenges that AI presents and said that the technology is also a huge opportunity.

"It's early days in AI and I think everyone uses it. On the same day, you're surprised at how good it is, and you're also surprised at how bad it is. It'll do things you think are hard, and then it'll fail to do some things you think are easy. I started my trip here in Hyderabad, which was a nice occasion because Microsoft is celebrating 25 years in the country, and that's been such a fantastic experience," Bill Gates said.

"So I was definitely challenging the team there. Hey, we have to make the accuracy better. And for now, it's like a Copilot. It helps suggest things. But the final decision about the medicine or the letter you want to write, the final decision, we still have to review that, even though it may make us a bit more creative and a bit more productive. You know, it seems like AI is a huge opportunity, but there are some challenges that come with that," Gates added.

Present "manipulated media" rules apply only to "videos that are created or altered by AI to make a person appear to say something they didn't say." Starting in February, the company began to add "Imagined with AI" labels to photorealistic images made with the Meta AI feature.