US researchers say computer programs known as bots were largely responsible for spreading a significant amount of misinformation on Twitter related to the 2016 US presidential election.
A research group from Indiana University published the findings in the science magazine Nature Communications.
The report says it analyzed 14-million messages spreading 400,000 articles on Twitter during 10 months in 2016 and 2017.
The group says it identified 6 percent of its studied Twitter accounts as bots, automated generators of tweets, and those accounts were linked to 31 percent of the low-credibility information.
The researchers say the likely bots spread the messages in the first few seconds after they were first published.
They also say a strategy often used by bots is to mention influential users in tweets that are linked to low-credibility content. They say, as an example, a single account mentioned President Donald Trump's account in 19 tweets, each linking to a same false claim.
Indiana University Professor Filippo Menczer says that introducing systems to prove that humans, not bots, are posting the messages would significantly curb misinformation from spreading on such social networks.