Page 10 - My FlipBook

fake news story that appeared on several sites. Still, the software
was not perfect. Some stories that later proved to be false were
not flagged. For example, a post about an NFL player suppos-
edly burning the American flag in his team’s locker room was not
flagged as fake even through it was proved to be false. To help
the software learn and improve as it works, the company manu-
ally updates its list of fake stories.
Social media companies like Facebook are also turning to AI to

fight fake news. In July 2018 Facebook announced that it was ac-
quiring the London-based start-up Bloomsbury AI. Currently, hu-
man workers oversee Facebook’s content. They inspect flagged
and reported posts and decide whether content violates the net-
work’s rules or whether it is fake news, hate speech, or other of-
fensive material. Facebook plans to increase its AI capabilities so
that algorithms perform more of this work, improving productivity
and allowing the platform to review more content, faster. Eventu-
ally, the company hopes to develop AI that can understand text,
images, and video well enough to effectively review content on
the platform. Bloomsbury AI’s team specializes in programming
computers to process and analyze large amounts of language
data, like that found on Facebook posts. On the Facebook Aca-
demics page, the company posted, “The Bloomsbury team has
built a leading expertise in machine reading and understanding
unstructured documents in natural language in order to answer
any question. Their expertise will strengthen Facebook’s efforts in
natural language processing research, and help us further under-
stand natural language and its applications.” 62

Government Efforts

While social media and technology companies are working to pre-
vent and remove fake news from their platforms, governments
around the world are also taking action. Many countries are con-
sidering or adopting new laws and regulations. A law passed in
France in 2018 gives French authorities the power to determine

59