TikTok begins working with a digital watermarking system called Content Credentials to help identify more photo, video, and audio-only content created using artificial intelligence tools from Microsoft, Adobe, OpenAI, and more .
The Chinese social networking giant, which is fighting a possible ban in the US citing national security concerns, is already labeling AI-generated content created using TikTok's AI effects tools. The company said the new move is part of its expanding efforts to combat misinformation and misinformation.
TikTok said in a statement: “AI enables great creative opportunities, but it can also confuse or mislead viewers if they don't know the content was generated by AI. “There is,” he said. “Labeling helps clarify that context.”
read more: How close to the truth is that photo? What you need to know in the age of AI
TikTok's move marks a new response from the tech industry to growing concerns about the proliferation of AI-generated content. Especially in the last few years, AI technology to create text, video, and audio has become much easier to use. (Check out his CNET hands-on reviews of AI image generation tools like Google's ImageFX, Adobe Firefly, and OpenAI's Dall-E 3, as well as other AI tips, explanations, and news on our AI Atlas resource page.)
At the same time, AI content has also become more trustworthy. Media and information experts warn that these converging trends could flood the internet with real lies and misinformation, posing serious risks to the public.
The concerns are not just theoretical. Earlier this year, a political consultant used AI to recreate US President Joe Biden's voice in a massive robocall. In that case, his AI recordings of hyper-realistic voices would have encouraged people in New Hampshire not to vote in the primary.
Experts believe this is just the beginning of the direction AI disinformation could go, especially for the 2024 presidential election.
TikTok isn't the only social media company using AI to identify posts. Last month, Meta, the owner of Facebook and Instagram, announced that it would mark videos, audio, and images as “Made with AI” if its system detects the involvement of AI, regardless of whether the creator discloses that information during the upload. announced plans to label it as “.
Google's YouTube also requires creators to disclose AI-manipulated videos, citing “realistic” portraits of people or scenes, or altered footage of real events or locations. OpenAI also announced that it will add AI identification data to all images generated using its system.
So far, users seem to broadly appreciate these efforts. Mehta found last month that 82% of more than 23,000 respondents from 13 countries supported AI-generated labels for content that “depicts people saying things they don't say.” He cited the survey results.
TikTok says it plans to expand its AI label efforts, including adding data to photos and videos created using AI tools on the platform, allowing people to download and share them elsewhere. Stated.
“As AI evolves, we continue to invest in combating harmful AI-generated content by evolving proactive detection models, consulting with experts, and partnering with industry peers on shared solutions. ” the company added.
read more: AI Atlas: A guide to today's artificial intelligence
Editor's note: CNET used an AI engine to create dozens of stories and label them accordingly. The notes you are reading are attached to articles that substantively cover the topic of AI, all written by professional editors and writers. Learn more about. AI policy.
