Twitter has detected in the last 15 days more than 1.5 million accounts globally suspected of manipulating or spreading spam (unsolicited messages that are usually advertising) in conversations related to SARS-Cov-2, the new coronavirus that has already It has infected more than a million people worldwide, according to sources from the company in Spain. In addition, the social network "has eliminated more than 1,100 tweets", including those of some relevant profiles, for spreading "misleading and potentially harmful content".
The action of Twitter, included within its policy of fighting disinformation, does not mean that it has deleted those 1.5 million accounts, but that it has "challenged" them. When machine learning (social network algorithms) detects "suspicious behavior", Twitter tries to confirm "whether or not there is a real person behind the account through a series of requests," reports a company spokeswoman. "We will not permanently suspend any account based solely on our automated systems," clarifies the social network, until the "human review" gives the go-ahead.
MORE TICKETS IN THIS BLOG
The unjustly nicknamed "1918 Spanish Flu"
Reporters Without Borders tracks how Chinese censorship helped spread the coronavirus
Who benefits from the avalanche of hoaxes about the coronavirus?
In the midst of the coronavirus crisis, Twitter automatically checks content first because most employees work remotely. This automated technology helps the company "review reports more efficiently" by selecting "the most likely content" that violates social network rules. But, as the company clarifies, "for content that requires additional context, such as misleading information about Covid-19," the intervention of "human teams" will be required. It also continues to monitor the accounts or content reported by users, although, as he warns, given the circumstances of confinement of employees, "the response time will be longer than normal."
Twitter has also expanded the concept of “definition of harm” on content related to the coronavirus and will prioritize the elimination of messages that “have a clear objective of a call to action that may present a direct risk to the health or well-being of persons". He cites as examples messages that deny health recommendations, advise harmful treatments or protective measures, deny scientific data about the infection, or lie about any type of information related to the coronavirus. For example, statements such as "Covid-19 does not infect children because we have not seen any cases of sick children" will be deleted. However, he makes it clear that he will not be able to take action with every tweet that contains “incomplete or disputed information about Covid-19,” in the context of “good faith” discussions or “ongoing studies related to potential medical interventions that they look promising. ”
Since the outbreak of the health crisis, Twitter has registered a notable increase in profiles. During the first quarter, it has recorded an average of daily users of "164 million", 23% more than in the first quarter of 2019 and 8% more than in the fourth of last year.
Follow us on Twitter and Flipboard