China News Service, San Francisco, February 6 (Xinhua) The American company Meta said on the 6th local time that it will begin to add tags to artificial intelligence-generated images uploaded to the company's social platforms Facebook, Instagram and Threads in the next few months.
Meta said in a statement that when users use Meta's artificial intelligence capabilities to create photorealistic images, the company takes several steps to ensure people know artificial intelligence is involved. These include adding visible tags to images, as well as embedding invisible tags and metadata in image files to help other platforms identify them.
Meta said that when posting realistic video or audio that is digitally created or altered, users will be required to mark the content with relevant tags, and violators may face penalties. If the Company determines that such content poses a particularly high risk of materially deceiving the public about an important issue, the Company may, at its discretion, mark it more prominently.
The company has been working with others in the industry to develop common standards for identifying AI-generated content, the statement said. Meta said it is developing tools to identify invisible tags at scale. When companies such as Google, OpenAI, Microsoft, Adobe, Midjourney and Shutterstock add metadata to images created by their respective tools, Meta can add tags to these images.
ABC reported that the risks posed by artificial intelligence-generated content have attracted widespread attention in recent weeks. Artificial intelligence-generated fake pornographic images of American pop singer Taylor Swift were spread on social platforms in late January, and the White House immediately called on Congress and technology companies to take action.
New Hampshire Attorney General John Formella said on January 22 that he would investigate a call made by a general artificial intelligence-generated voice of U.S. President Joe Biden to voters, which called on voters not to participate in the party's primary election for the presidential election. Nick Clegg, president of global affairs at Meta, said that as Americans go to the polls in 2024, technology companies should take action to ensure that users can identify whether online content is authentic. (over)