- Tram Ho
Facebook said in a statement today that deepfake videos and staged media content will be banned from its social networking platform.
The company said it is trying various approaches to solve the problem, including investigating shady, deceptive behavior, in AI-generated content, and working with academia. , government, and industry companies to more accurately identify staged content.
” The staging content can be done through simple technologies like Photoshop, or through sophisticated tools that use artificial intelligence or ‘deep learning’ techniques to create distorted videos. reality – often called ‘deepfake’ “- Facebook said. ” While such videos are still rare on the Internet, they present a significant challenge to technology and society as they are increasingly taken advantage of.”
The company defines “deepfake” content types to be removed from its platform as follows:
– Content that has been edited or aggregated – except for refinements to increase transparency or display quality – in ways that are unclear to a normal person and will likely mislead someone , making them think that a subject in the video says things they don’t really say.
– Content is a product of artificial intelligence or machine learning, which involves stitching, replacing, or overlaying other content into videos, making it seem real.
Facebook’s policy does not affect content that is staged for humor or satire – the company added.
Videos marked as infringing will undergo a review by third-party testing partners to determine if the video content is truly false. If so, Facebook will seek to ” significantly reduce its distribution” on newsfeeds or refuse to publish it, if it is used as an advertisement.
And yet, those who seek to share content before it is completely removed from the site will receive a warning that the content is not true.
The Facebook boss also became a victim of Deepfake
Facebook added that it has established a partnership with Reuters to ” help news centers around the world identify deepfake and staged content through a free online training course”.
“News organizations are increasingly relying on third parties to review a large number of images and videos, and identifying staged content is a significant challenge. This is aimed at supporting news centers to do this “- Facebook stated.
In September of this year, Facebook Chief Technology Officer Mike Schroepfer said the company is creating its own deepfake content to help better detect staged content and remove it.
” The goal of this is to create technology that everyone can use to better detect when AI is used to change the content of a video to mislead viewers,” Schroepfer wrote in an article. blog.
Last year, the company faced a backlash from members of Congress, including Alexandrian delegate Ocasio-Cortez and presidential candidate, Senator Elizabeth Warren, because of a policy on auditing. Look up information in political advertisements.
Facebook says that in most cases it will not check for political advertisements to protect the right to freedom of speech. This controversial policy was criticized after President Donald Trump ran political ads with false information about former Vice President Joe Biden.
” Posts and advertisements from politicians are generally not checked” – according to Facebook’s policy. ” To evaluate when this should be done, we will ask our test partners to review information about politicians at all levels.”
Source : Genk