The San Francisco social-media company currently attaches small blue notices to false tweets.
It is assessing how to make these signals more "overt" and "direct," Twitter's Yoel Roth told Reuters.
Roth made no mention of whether the changes would be implemented before the Nov. 3 U.S. election.
The changes will include testing a reddish-magenta color that is more visible, Roth told the news service.
Twitter reduces the reach of tweets that it labels for false content by limiting their visibility and not recommending them in search results, Reuters reported.
Feedback from users tells the company that they want to know whether an account has been repeatedly labeled, Roth said. Twitter will consider whether to flag users who constantly post false information, he said.
Twitter said it had labeled thousands of posts, including some tweets by President Donald Trump.
Since first labeling Trump in May, Twitter put gray warning overlays over 10 of the president’s tweets.
In May, Twitter began labeling fabricated media, expanding its labels to coronavirus misinformation, misleading tweets about elections, and civic processes.
Twitter has been criticized for its transparency regarding its interventions, according to Reuters. The company doesn’t keep public lists of when it applied labels or disclose data that would allow outsiders to assess how those labels affect the spread of a tweet or its reactions.
The company consults with partners, including election officials, on its labeling.
In September, the social-media company said it would label or remove tweets claiming election victory before the results were confirmed.