Twitter founder: AI will remain integral to fight against extremism

The social media giant has come under fire for not doing enough to de-platform extremists

Biz Stone addresses a summit in Abu Dhabi in 2010. The National

Artificial intelligence and behavioural data will continue to be the best tools to stop violent extremism on social media in its tracks before it is too late, the founder of Twitter has said.

Responding to a question from The National at the One Young World Summit on Friday, Biz Stone, who co-founded the social network alongside Jack Dorsey in 2006, said that the rise of extremism planned online was something "the industry never saw coming" when social media websites first came about.

“We weren’t thinking – oh there might be state sponsored, concentrated co-operative agent working and amplify an argument that is already happening, and hyper-polarise it to the point where it turns into extremism. We just didn't think that and it wasn't in our heads so we had to play catch up,” he said.

Pressure has mounted on social media giants to quickly remove extremist and dangerous content, after terrorists in attacks earlier this year - including the Christchurch Attack in New Zealand and the El Paso Walmart attack - used social networks to incite hatred and broadcast their attacks.

Twitter meanwhile, has been criticised for not devoting as many resources to removing, or “de-platforming,” white supremacist accounts as it has to ISIS accounts.

Mr Stone said Twitter previously used to be “more reactive than proactive”, relying more on human personnel to track down extremist content and hate crime, but the platform is now “catching up”.

“Now what we can do is realise that's not the future  - we just cant look at a billion tweets a day that are in Farsi, Japanese and everything else. Where we are moving to, is looking at behaviour instead of content because most of the time certain behaviour is linked to certain kinds of content.”

He said that since Twitter adopted AI to crack down on extremist content, it has had a profound effect and the company now captures around 40 per cent of all dangerous content immediately.

Because of AI and algorithms, Twitter can now suspend accounts that harass users immediately, whether a bot or a real person, he said.

“I can only imagine that it’ll get more and more sophisticated as we get more and more data of people's behaviours. But not just individual behaviour but also group behaviour - we need start being more proactive.”

He added that most of Twitter’s acquisitions have been in AI recently to help the social networking site better identify extremist content and remove it immediately.