“We’re testing a feature for you to report Tweets that seem misleading – as you see them,” the social network said from its safety and security account.

From yesterday, a button would be visible to some users from the United States, South Korea and Australia to choose “it’s misleading” after clicking “report tweet”.

Users can then be more specific, flagging the misleading tweet as potentially containing misinformation about “health”, “politics” and “other”.

“We’re assessing if this is an effective approach so we’re starting small,” the company said.

“We may not take action on and cannot respond to each report in the experiment, but your input will help us identify trends so we can improve the speed and scale of our broader misinformation work.”

Twitter, like Facebook and YouTube, regularly comes under fire from critics who say it does not do enough to fight the spread of misinformation. But the platform does not have the resources of its Silicon Valley neighbours, and so often relies on experimental techniques that are less expensive than recruiting armies of moderators.

Moderators are ultimately responsible for determining which content actually violates Twitter’s terms of use, but the network has said it hopes to eventually use a system that relies on both human and automated analyses to detect suspicious posts.