Are we over worrying fiddling with polls?
I've seen a lot of press recently about the actions of a Nation State attempting to interfere in the outcome of an election in another Nation State. My immediate thought is so what? Any Nation State should be concerned with the politics of other Nation States and ought to be sufficiently concerned to make the result as favourable as possible to them. This will not be done overtly - you will not find a Nation State saying we want Bob to win and not Alice. Rather I'd expect every Nation State would have discussions at government level on how they would work with any other Nation State in the event of change of government or even in the event of change of a senior minister.
So in the age of widespread social media, of global media, of global business supply chains, of interlinked economies, the nature of influence can be very nuanced. So lets put some names here: Would Russia be expected to want to influence the votes in either the US Presidential election, or the Brexit referendum? I'd say the answer is obviously yes.
So what's the controversy? I think it's technology allowing old fashioned espionage to get a massive boost. The shady worlds of Graham Greene, and of John le Carré's Smiley, where a word here, a bit of pressure there, and simple coercion peppered around, could swing a situation in the favour of one side or the other. So what happens when millions of words are spread around that cast doubt on one party and subtly promote the other? What if those words are coordinated to swing a particular position? What if small acts of pressure are multiplied at the speed of the internet? What if the source of all of this can be masked by the very nature of global economics? Then a Nation State would be naive to not take advantage. So if we accept that this is legitimate in the way in which Nation States keep tabs on each other then how does the victimised Nation State protect itself?
The first step in defence is to recognise that this will happen. However defence is difficult if the intent is to swing a government by a few seats to tilt the balance of power. Elections in first past the post results can be seen to swing from party to party in fairly predictable patterns with a period of status quo often sitting between swings. So in the UK we've gone from Tory to Labour to Tory to Labour in a long series of cycles. It took a fairly radical societal change to move from Whigs and Tories to Labour and Tories and every so often the 3rd party comes close to tipping the balance. How many votes are needed to swing to change political direction? Can a marketing and data analysis of trends find the critical lever? That is the promise of big data and it'd be a foolish person that would suggest we cannot be influenced. If we cannot be influenced then advertising would not work and quite simply there is lots of apocryphal evidence that advertising works.
Can we identify good versus malicious interference? This is always going to be a difficult thing to determine. If I am a fan of Alice and I hear bad things about Bob they will probably reinforce my support of Alice. But. If I have no strong opinion of who to choose between Alice and Bob but have strong opinions on some topic then hearing that (say) Bob is strongly at odds with my view whilst Alice is much more aligned to my view then maybe I will swing towards Alice. Played over a wide enough playing field with enough subtlety a whole raft of swing voters could be edged in one direction or another. Party politicians already do this and they employ marketing agencies to help them. They play the media in this too, and they'll have overt supporter from sectors of the media. So when does the support and campaigning become a matter for national security concerns? I guess the answer is when a foreign nation state is seen to be the sponsor.
Ok - we can identify some input to an election being sourced in a foreign nation state, we may be able to trace the money as coming from that foreign nation state. What can be done about it? My guess is that any countermeasure requires a lot of intelligent analysis to sort wheat from chaff. The CIA paradigm of cyber-security doesn’t really have a comfortable fit to this kind of problem but this is the very core of cyber security problems. We appear to want a defence against undue influence and whilst protection of a nation state is obviously a leading contender for the application of such a defence there are many others that will benefit. For example we can counter sextortion, radicalisation, net-borne viruses and others.
So in the age of widespread social media, of global media, of global business supply chains, of interlinked economies, the nature of influence can be very nuanced. So lets put some names here: Would Russia be expected to want to influence the votes in either the US Presidential election, or the Brexit referendum? I'd say the answer is obviously yes.
So what's the controversy? I think it's technology allowing old fashioned espionage to get a massive boost. The shady worlds of Graham Greene, and of John le Carré's Smiley, where a word here, a bit of pressure there, and simple coercion peppered around, could swing a situation in the favour of one side or the other. So what happens when millions of words are spread around that cast doubt on one party and subtly promote the other? What if those words are coordinated to swing a particular position? What if small acts of pressure are multiplied at the speed of the internet? What if the source of all of this can be masked by the very nature of global economics? Then a Nation State would be naive to not take advantage. So if we accept that this is legitimate in the way in which Nation States keep tabs on each other then how does the victimised Nation State protect itself?
The first step in defence is to recognise that this will happen. However defence is difficult if the intent is to swing a government by a few seats to tilt the balance of power. Elections in first past the post results can be seen to swing from party to party in fairly predictable patterns with a period of status quo often sitting between swings. So in the UK we've gone from Tory to Labour to Tory to Labour in a long series of cycles. It took a fairly radical societal change to move from Whigs and Tories to Labour and Tories and every so often the 3rd party comes close to tipping the balance. How many votes are needed to swing to change political direction? Can a marketing and data analysis of trends find the critical lever? That is the promise of big data and it'd be a foolish person that would suggest we cannot be influenced. If we cannot be influenced then advertising would not work and quite simply there is lots of apocryphal evidence that advertising works.
Can we identify good versus malicious interference? This is always going to be a difficult thing to determine. If I am a fan of Alice and I hear bad things about Bob they will probably reinforce my support of Alice. But. If I have no strong opinion of who to choose between Alice and Bob but have strong opinions on some topic then hearing that (say) Bob is strongly at odds with my view whilst Alice is much more aligned to my view then maybe I will swing towards Alice. Played over a wide enough playing field with enough subtlety a whole raft of swing voters could be edged in one direction or another. Party politicians already do this and they employ marketing agencies to help them. They play the media in this too, and they'll have overt supporter from sectors of the media. So when does the support and campaigning become a matter for national security concerns? I guess the answer is when a foreign nation state is seen to be the sponsor.
Ok - we can identify some input to an election being sourced in a foreign nation state, we may be able to trace the money as coming from that foreign nation state. What can be done about it? My guess is that any countermeasure requires a lot of intelligent analysis to sort wheat from chaff. The CIA paradigm of cyber-security doesn’t really have a comfortable fit to this kind of problem but this is the very core of cyber security problems. We appear to want a defence against undue influence and whilst protection of a nation state is obviously a leading contender for the application of such a defence there are many others that will benefit. For example we can counter sextortion, radicalisation, net-borne viruses and others.
My challenge is to work across the community in standards to
evolve the CIA paradigm in such a way that we can give leverage to the
intelligent human that can make a targeted prosecution against the “bad party”
in these kind of things.
Comments
Post a Comment