Twitter is taking another crack at stamping out trolls and abusers

Jack Dorsey, CEO of Twitter and Square.
Rebecca Cook | Reuters

Twitter is launching new safety features to make it easier for users to avoid trolls.

The social media platform said Tuesday it will stop so-called trolls from creating new abusive accounts and will collapse "low-quality" and potentially abusive tweets from feeds and searches.

That means Twitter will try to identify people who have been permanently suspended and block them from creating new accounts. The "safe search" will remove sensitive content and blocked and muted accounts from search results. Tweets that Twitter deems "potentially abusive" would need to be specifically sought out.

The changes come amid widespread criticism of Twitter, which BuzzFeed News called "honeypot" for harassment. BuzzFeed's reporting revealed that Twitter staff struggled to balance the promotion of free speech while curbing bullying.

Leslie Jones of "Saturday Night Live" made a temporary, high-profile exit from Twitter last year after being targeted by trollers. Salesforce CEO Marc Benioff ruled out a bid to purchase Twitter in part because of concern about its reputation for handling online abuse. Even former CEO Dick Costolo has said that Twitter "sucks" at dealing with bullying.

Twitter has modified its product, introducing improved ways to report abusive tweets. But Twitter's vice president of engineering, Ed Ho, tweeted last month that the company "didn't move fast enough" last year to make Twitter safer.

"With every change, we'll learn, iterate, and continue to move at this speed until we've made a significant impact that people can feel," Ho wrote in announcing the changes.

— CNBC's Julia Boorstin and Berkeley Lovelace Jr. contributed to this report.