Twitter Inc. released a new transparency report Tuesday that reveals how busy it has been weeding out extremism on its platform.

The report follows mounting pressure from governments for social media platform to get tougher on terror. The company said that it had removed 299,649 accounts that promoted terrorism or endorsed violence enacted by terrorist organizations in the first half of 2017. Three-quarters of those accounts were suspended before they even posted their first tweet, but Twitter said the number of removed accounts was still a 20 percent decline from the previous six months.

Twitter also said 95 percent of the accounts removed were a result of its own internal automation tools. “Our anti-spam tools are getting faster, more efficient and smarter in how we take down accounts that violate our policy,” Twitter said in a statement. Fewer than 1 percent of the suspensions came as a result of requests from authorities.

In the previous transparency report, Twitter wrote that just 74 percent of account suspensions were a result of its automation tools. Twitter hasn’t said how those tools work, except that anything breaching the company’s terms of service gets flagged. In a statement, Twitter explained, “We are reluctant to share details of how these tools work as we do not want to provide information that could be used to try to avoid detection.”

Some 935,897 accounts were suspended from August 2015 to June 2017. In the three months prior to that, Twitter said, that it was averaging 328 million active users monthly.

Twitter and other large tech firms, notably Facebook Inc. and Google LLC, have been under considerable pressure following numerous reports that their services are continually being used to disseminate extreme ideology and recruit new members to terrorist organizations.

Following the attacks in London this year, British Prime Minister Teresa May issued the damning statement, “We cannot allow this ideology the safe space it needs to breed. Yet that is precisely what the internet, and the big companies that provide internet-based services, provide.”

Like many pundits have pointed out, it is an uphill battle given the mind-boggling amount of content that comes through large social networks daily. On top of that, free speech advocates worry that moderation could turn to an oppressive form of censorship and automation tools could end up suppressing needed debate. Other experts have said blaming the Internet and going down the road of censorship might be a dangerous and hopeless path to follow.

Image: Andreas Eldh via Flickr