Baidu Invites China’s Cybercops to Label, Rebut Fake News

(Bloomberg) -- Baidu Inc. is building a system to allow China's cybercops to spot and fix “online rumors” deemed a threat to stability, allowing police agencies to insert themselves directly into everything from its search results to discussion forums.

The platform links 372 police agencies who will use sophisticated artificial intelligence-driven tools to monitor and respond to fake news, blogposts and other items across about a dozen Baidu services, including the popular search engine, the official Xinhua News Agency reported. More than 600 organizations and experts in different areas will be enlisted to weigh in on their respective fields, according to an email sent by Baidu. They included official organs such as the Chinese Academy of Social Sciences, as well as media outfits such as Shanghai United Media Group and Caijing.

Internet giants from Facebook Inc. to Twitter Inc. are struggling to deal with a proliferation of spurious news articles across social media services. Baidu's approach allows the Chinese government to intervene directly and write articles in rebuttal. Items that its system decides are fake will be clearly labeled a “rumor” at the very top of search results, alongside an explanation penned by the relevant agency or organization, according to a sample page Baidu provided.

The same system will be employed across products from its news aggregator and online forums to Quora-like Q&A service, Baidu spokeswoman Whitney Yan said. 

The company's announcement comes days after Chinese cyberspace regulators upbraided and fined Baidu, Tencent Holdings Ltd. and Twitter-like site Weibo for broadcasting pornography, violent content and fake news. All three have said they will cooperate and remove objectionable material.

China is tightening scrutiny over domestic internet content in the run-up to an important Communist Party congress that's expected to consolidate President Xi Jinping's authority. Intent on muzzling potential sources of disruptive information, the government has shut livestreaming services and websites, tightened regulations governing VPNs, and issued repeated warnings about the need to clean up content through various agencies.

On Wednesday, the CAC also warned Facebook Inc.'s WhatsApp against carrying “illegal content,” after the messaging service experienced a series of intermittent outages.

Welcome to Xi's Net: Where Politics, Porn and Pooh Are Forbidden

To contact Bloomberg News staff for this story: Gao Yuan in Beijing at ygao199@bloomberg.net.

To contact the editors responsible for this story: Peter Elstrom at pelstrom@bloomberg.net, Edwin Chan, Reed Stevenson

With assistance from Gao Yuan