BEFORE LOOKING AT how Ireland handles social media in the upcoming referendum to repeal the Eighth Amendment to the Constitution we should first seek to examine how other countries – particularly the US – have experienced its (mis)use, and if possible see what can be learned.

2017 was a year where the big social platforms went from a position of outright denial that their platforms had been used by foreign actors during the United States election, to a position were they all but admitted that not only did it happen, but that it was a far larger operation than they realised.

Last Friday, Facebook and Twitter both made important announcements in relation their respective roles in relation to election processes.

For its part, Twitter published an update on its investigation into Russian interference in the 2016 US Presidential Election. Twitter revealed that it was “emailing notifications to 677,775 people in the US who followed one of these accounts or retweeted or liked a Tweet from these accounts during the election period”.

This might sound like an exercise in transparency but when you dig deeper you realise Twitter is leaving lots out.

Twitter is only including accounts believed to be connected to the Russian Internet Research Agency (IRA) – not other possible troll farms or constructed bot armies; they also only include the 10 weeks leading up to the election, not the period from the beginning of the primaries; they also leave out people like us – who live outside the US but followed the US election closely.

‘Russian actors created 80,000 Facebook posts that reached 126 million people in the US’

Facebook meanwhile has been dragged kicking and screaming to an admission that its platform can be used to manipulate elections. Facebook CEO Mark Zuckerburg in 2016 laughed off a suggestion that fake content could have influenced the US election. “I think it’s a pretty crazy idea” that it “influenced the (US) election in any way,” he said.

Many at the time thought it was crazy that Zuckerburg either didn’t understand how Facebook was being used, or was willfully negligent.

Fast forward to a statement Facebook published just this week: “Although we didn’t know it at the time, we discovered that these Russian actors created 80,000 posts that reached around 126 million people in the US over a two-year period.”

Not only was there an influence operation, but it likely reached a majority of US voters. The statement continued:

At its best, [Facebook] allows us to express ourselves and take action. At its worst, it allows people to spread misinformation and corrode democracy.

Facebook has announced a number of measures that are currently being trialled to increase transparency on the platform – but it seems unlikely these measures will be rolled out in time for the referendum here.

And we haven’t even touched on how echo chambers and the spread of false information across the network can undermine how people see the world, and reality, or on how similar behaviours are seen on other platforms like Google and YouTube.

How do we think about this issue in relation to the upcoming referendum in Ireland?
There are a number of overlapping problems related to how a referendum should be run freely and fairly in light of the revelations of the past year. And we should look at this problem from first principles and then see whether our laws and rules around elections and referenda are adequate enough to ensure they are free and fair.

The first lesson to learn from the US election is that attempts to message voters by Russia started long before election day – there was preparation of the pages, the ads and the building of communities around divisive topic long before polling day.

This leads to some obvious questions for legislators: Should foreign actors be allowed to do this in the context of an upcoming referendum? For both domestic and foreign actor purposes when does the clock start ticking on political messaging in the context of the referendum in question? One month before the vote date? Six months? 12 months? Should dates be announced very well in advance?

A second lesson from the US election was the use of bots and/or “cyborgs”. Bots are automatically created accounts that tend to tweet automatically or semi-automatically, while cyborgs are networks of Twitter accounts run manually by one or more people often in order to give the impression that a fringe view is more mainstream than it actually is.

Should anyone be able to create false impressions of reality?

Ireland only has a population of 4 million people. I could conceivably create 500,000 bot accounts tomorrow – with the right tools and determination – I could then stock these bot accounts with divisive rhetoric on the subject of abortion – taking one side, or the other, or both – and then programme these bots to look like real people and tweet with the hashtags of all the major TV and radio shows in Ireland. Theoretically I could do the same to the comments below TheJournal’s stories – including this one.

Any Twitter users who follow those hashtags would see the result of my one-man bot army and be duped into thinking the views I express via this accounts are popular views – when in fact it’s just me. No law prohibits this behaviour in the context of an election or referendum. Should a law prohibit it? Should anyone be able to create false impressions of reality using manufactured armies of puppet Twitter and Facebook accounts?

A third lesson from the US election was the use of paid ads. Here is where things get more complicated. But let’s start with the basic question: whether I’m pro-repeal or anti-repeal, and I’m a high net worth individual and I want to spend €1 million on ads right now targeting Irish women between the ages of 30 and 60 on Facebook, imploring them, say, to stop the repeal of the amendment – is that my right, or is my right restricted because money should not be the defining factor in how the people decide this issue?

SIPO in its guidance notes states: “As the legislation stands, a person, which could include an individual or an organisation, who does not receive a donation and who proposes to incur expenditure at a referendum is not required to register at all with the Standards Commission”.

Who gets a say in our electoral processes? 

This leads to further questions: Should there be a permanent referendum commission? An electoral commission? To what extent should actors, both domestic and foreign, be allowed to engage in our electoral processes via social media and should we limit it, and if so to what extent, and how would it be regulated?

There are no easy answers to these question but there are steps we can take to mitigate and some of the recently proposed amendments by Kildare TD James Lawless have merit. As do other reforms proposed in the US by Klobuchar, Warner and McCain.

Ireland is now at fundamental risk of our democratic processes being influenced or manipulated via the vector of social platforms, either by using money or by gaming the system. We must ask ourselves who gets a say in our electoral processes, how they have say, and how that process is regulated. We must immediately learn the lessons of other countries – a more pressing issue for us given the sheer number of referenda planned.

In light of events in 2016, and in France and Germany in 2017 and during the Brexit referendum in the UK, what is now urgently needed is a fundamental review of all electoral law that bring us a modern legal framework, with enforcement mechanisms designed to suit the world we live in today, not the world we lived in 20 or 30 years ago.

If we don’t, as Facebook stated, we risk the further spread of mis or disinformation and the corrosion of our democracy.

Gavin Sheridan is a transparency campaigner and co-founder of legal intelligence platform Vizlegal.