Twitter, which has been under pressure to do more to combat harassment, said Tuesday that it's being more proactive about flagging tweets that violate its rules against hate speech, threats and other abusive content.
The social media company said that it used to only review abusive tweets that users reported to the company. Now it's using technology to identify abusive tweets before it receives a report.
"People who don't feel safe on Twitter shouldn't be burdened to report abuse to us," Donald Hicks, head of Twitter's services, and David Gasca, Twitter's product lead on health, said in a blog post.
About 38% of content that violates Twitter's rules was flagged by the company, which said it used technology to find these tweets. That's up from zero percent during the same period last year. Twitter has lagged behind other social networks such as Facebook, which have also been using artificial intelligence to identify offensive content before a user reports it.
Twitter has been criticized for being a "toxic place," especially for women. Last year, a study by Amnesty International and Element AI found that received "abusive" or "problematic" tweets every 30 seconds on average.
One of those journalists grilled Twitter CEO and co-founder Jack Dorsey on Tuesday, while he took questions via Twitter during an onstage interview at the TED2019 conference in Vancouver.
"I'd like to know why a video that showed me being beaten up & threatened with a gun to soundtrack of Russian anthem stayed up for 72 hours despite 1000s of complaints," investigative journalist Carole Cadwalladr, who writes for The Guardian and delivered a talk Monday at the conference, demanded of Dorsey in a tweet.
Dorsey didn't answer Cadwalladr's question, but he acknowledged that the company still has a lot of work to do to make Twitter a healthier place.
"We've seen abuse, we've seen harassment, we've seen manipulation, automatic and human coordination, misinformation," he said during the TED conference. "What worries me most is our ability to address it in a systemic way that is scalable."
Twitter's announcement also came a day after BuzzFeed reported that Twitter left up death threats against Rep. Ilhan Omar, a Minnesota Democrat and one of the first two Muslim women elected to Congress, so law enforcement could investigate them. At the same time, Twitter has also faced allegations that it suppresses conservative speech. During a last week, the company again denied those assertions.
If he had to start Twitter again, Dorsey said during the TED interview, he wouldn't emphasize "likes" and follower counts as much.
"I don't think I would even create 'like' in the first place -- because it doesn't actually push what we believe now to be the most important thing, which is healthy contribution back to the network," he said.
Twitter may work better if it becomes an "interest-based network," Dorsey said, where people are encouraged to follow interests rather than individual accounts.
Twitter, which has 321 million monthly active users, also said it's cracking down on users who create new accounts after they've been suspended. The company said it suspended 100,000 accounts between January and March for creating new accounts after a suspension, a 45% increase compared with the same period last year.
Twitter said it'll be working on new changes in the coming months, including adding a notice if a tweet violates the company's rules but stays up because it's in the public interest.
Twitter also said it'll launch a feature in June that lets people. The social network acknowledged it was developing the option back in February, saying it would help users protect their conversations.
Originally published April 16, 2:39 p.m. PT
Update, 2:40 p.m.: Adds Twitter CEO's remarks at TED, and more background.
Update, April 17: Adds details about the hide replies feature.