China’s state-run Xinhua News Agency debuts ‘AI anchor’ to read the news

[Presence-evoking technology is impacting the delivery of news, raising important ethical questions. This story is from Mashable, where it includes a 0:31 minute video of an “AI anchor”; for more information and a longer video, see coverage in The Verge. –Matthew]

[Image: Source: Press From]

China’s state-run Xinhua News Agency debuts ‘AI anchor’ to read the news

By Matt Binder
November 8, 2018

The state-run press agency in China, Xinhua, will now deliver news using “AI anchors” made of digital composites that use synthesized voices to “read” the news.

These anchors look realistic, but are actually digital composites that are rendered using actual footage of human anchors reading the news. The AI anchors are able to “read” any text fed into the artificial intelligence using synthetic voices that are created through the use of composite audio recorded from the real-life anchors.

According to the South China Morning Post, an English-language anchor and a Chinese-language anchor were created by Xinhua in partnership with local Beijing search engine company Sogou. The anchors premiered on Wednesday at the World Internet Conference in China.

Xinhua News Agency says that the AI anchor, “can work 24 hours a day on its official website and various social media platforms, reducing news production costs and improving efficiency.” The state-run news agency considers the AI anchors to be members of its reporting team.

Xinhua didn’t divulge any additional information into the technology being used to create the AI anchors, but the final product feels very reminiscent of the “deep fakes” videos that emerged in the West.

Deep fakes are essentially “face-swapped” videos created by artificial intelligence that scans hundreds of still frames from a video and then uses those analyzed frames to create new, manipulated video that looks like the real thing. Amateur hackers have used this type of software to create fake celebrity pornography.

Earlier this year, sites like Reddit were inundated with digitally altered video that convincingly superimposed Hollywood stars’ faces onto adult film actresses’ bodies. Another emerging example of its use case is for fake news. In April, Buzzfeed produced a fake video of President Obama (voiced by comedian Jordan Peele) that was created using similar technology. The video was meant to highlight how digitally manipulated video, powered by AI, is becoming more accessible, and how it may be used in an era increasingly filled with fake news, misinformation, and propaganda.

The U.S. Defense Department is concerned enough about people believing fake AI-created imagery to have already started working on tools to combat it.

Seeing equivalent — if not even more advanced — AI technology being used to quite literally power a news anchor, one run at a state-run news agency in China no less, should give us all pause.

This entry was posted in Presence in the News. Bookmark the permalink. Trackbacks are closed, but you can post a comment.

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

*
*

  • Find Researchers

    Use the links below to find researchers listed alphabetically by the first letter of their last name.

    A | B | C | D | E | F| G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z