Twitter has agreed to hand to Australian authorities the account names and IP addresses of users suspected of bullying or other harmful tweets, according to communications minister Stephen Conroy.
Senator Conroy told the Today breakfast show on Thursday morning that the agreement came out of a meeting between the Department of Broadband, Communications and the Digital Economy, the Australian Federal Police and Twitter's head of global public policy, Colin Crowell.
A spokesman for Senator Conroy said the specifics of the agreement would be expanded in future discussions but that metadata could include information such as account names, the time a tweet was posted and the IP address of the Twitter user; all information already kept by the social network.
"What they've [Twitter] agreed to do, is firstly they'll have a direct line of communication — something that wasn't there before. Secondly, if the federal police get in touch and say, 'Look we've got imminent threats being made, possible danger, serious threats', then they'll preserve the data, they'll cooperate with the police," Conroy told the show.
The social network would keep "key information" for Australian police until legal processes around a particular user or set of tweets was completed.
"We've got to send a message and Twitter are now onboard to send the message that this sort of behaviour is not acceptable, that you can't hide behind anonymity," Conroy said.
"A lot of the trolling that goes on is because people think, 'They'll never find out who I am and so I can behave this way'. This is an important message."
Despite Twitter's reported reassurances, the social network insists on only handing over user information requests from non-US authorities through US court-ordered subpoenas — a measure that has proved more successful in recent weeks — or mutual assistance treaties.
"If Twitter is contacted directly by law enforcement, we can work with them and provide assistance for their investigation as well as guidance around possible options," the company said in its abusive behaviour policy.
Though Twitter has moved to align itself more closely with specific country guidelines and requests from respective authorities, it has maintained that any actions are purely reactive.
It reassured users earlier this year that proactive measures such as intervening in disputes, or filtering tweets for potentially abusive material was "neither desirable nor realistic".
However, the move by Australian authorities, and particularly the AFP, to proactively investigate instances of bullying or abuse on Twitter partially come as an about-turn for most agencies who have, to date, refused to acknowledge any interaction with the social network.
Twitter reported receiving less than ten user information requests from Australian authorities in the first six months of the year — compared to 679 requests from US authorities in the same time frame — of which the social network complied a third of the time.
Media Watch reported on Monday that the AFP had relegated such harrassment as a state matter.
However, Tasmanian police recently urged citizens to stop alerting it to every "abusive or harassing" comment posted to Facebook or other social media sites, claiming it was "not the role of Tasmania Police to censor internet content".
iTnews asked AFP about whether it had previously asked for such information from Twitter, or whether it would deem the IP addresses — which may not directly correlate to a specific user or computer — to be useful, but did not receive a response by the time of publication.