The government and tech companies can't prevent 'fake news' during the election – only the public can
- Written by Michael Jensen, Senior Research Fellow, Institute for Governance and Policy Analysis, University of Canberra
We’re only days into the federal election campaign and already the first instances of “fake news” have surfaced online.
Over the weekend, Labor demanded that Facebook remove posts it says are “fake news” about the party’s plans to introduce a “death tax” on inheritances. Labor also called on the Coalition to publicly disavow the misinformation campaign.
An inauthentic tweet purportedly sent from the account of Australian Council of Trade Unions secretary Sally McManus also made the rounds, claiming that she, too, supported a “death tax”. It was retweeted many times – including by Sky News commentator and former Liberal MP Gary Hardgrave – before McManus put out a statement saying the tweet had been fabricated.
What the government and tech companies are doing
In the wake of the cyber-attacks on the 2016 US presidential election, the Australian government began taking seriously the threat that “fake news” and online misinformation campaigns could be used to try to disrupt our elections.
Last year, a taskforce was set up to try to protect the upcoming federal election from foreign interference, bringing together teams from Home Affairs, the Department of Finance, the Australian Electoral Commission (AEC), the Australian Federal Police (AFP) and the Australian Security Intelligence Organisation (ASIO).
The AEC also created a framework with Twitter and Facebook to remove content deemed to be in violation of Australian election laws. It also launched an aggressive campaign to encourage voters to “stop and consider” the sources of information they consume online.
For their part, Facebook and Twitter rolled out new features aimed specifically at safeguarding the Australian election. Facebook announced it would ban foreign advertising in the run-up to the election and launch a fact-checking partnership to vet the accuracy of information being spread on the platform. However, Facebook will not be implementing requirements that users wishing to post ads verify their locations until after the election.
Twitter also implemented new rules requiring that all political ads be labelled to show who sponsored them and those sending the tweets to prove they are located in Australia.
While these moves are all a good start, they are unlikely to be successful in stemming the flow of manipulative content as election day grows closer.
Holes in the system
First, a foreign entity intent on manipulating the election can get around address verification rules by partnering with domestic actors to promote paid advertising on Facebook and Twitter. Furthermore, Russia’s intervention in the US election showed that “troll” or “sockpuppet” accounts, as well as botnets, can easily spread fake news content and hyperlinks in the absence of a paid promotion strategy.
Facebook has also implemented measures that actually reduce transparency in its advertising. To examine how political advertising works on the platform, ProPublica built a browser plugin last year to collect Facebook ads and show which demographic groups they were targeting. Facebook responded by blocking the plugin. The platform’s own ad library, while expansive, also does not include any of the targeting data that ProPublica had made public.
Read more: Russian trolls targeted Australian voters on Twitter via #auspol and #MH17
A second limitation faced by the AEC, social media companies, and government agencies is timing. The framework set up last year by the AEC to address content in possible violation of electoral rules has proven too slow to be effective. First, the AEC needs to be alerted to questionable content. Then, it will try to contact whoever posted it, and if it can’t, the matter is escalated to Facebook. This means that days can pass before the material is addressed.
Last year, for instance, when the AEC contacted Facebook about sponsored posts attacking left-wing parties from a group called Hands Off Our Democracy, it took Facebook more than a month to respond. By then, the group’s Facebook page had disappeared.
Authors: Michael Jensen, Senior Research Fellow, Institute for Governance and Policy Analysis, University of Canberra