First, I feel obligated to reference the following quote, as I do when I teach someone how to pick a lock:
If a lock, let it have been made in whatever country, or by whatever maker, is not so inviolable as it has hitherto been deemed to be, surely it is to the interest of honest persons to know this fact, because the dishonest are tolerably certain to apply the knowledge practically; and the spread of the knowledge is necessary to give fair play to those who might suffer by ignorance.
-- From A. C. Hobbs (Charles Tomlinson, ed.), Locks and Safes: The Construction of Locks. Published by Virtue & Co., London, 1853 (revised 1868).
Second, I feel obligated to say this is not serious advice, or intended as an operating manual, but it's intended to demonstrate how easy it would be to manipulate the Canadian elections this year.
Third, while I am associated with a political party, I would go under oath stating that I have not participated in any of these steps for them, or myself as a candidate, even knowing that I could.
OK so here we go. The biggest loophole to winning the 2019 Canadian elections is to realize that Canadian political parties are not bound by Canadian privacy laws. This is important, as laws like PIPEDA that govern the collection, use, and disclosure, of Canadians personal information, is governed by the Canadian federal privacy commissioner.
As an aside, this law isn't really respected by the private sector today even though it is regulated, as the commissioner doesn't have order making power even if you are guilty of violating privacy laws -- and most relevant, there are no repercussions of consequence. Things are even worse in the public sector in terms of lack of repercussions, but let's get back to the political arena for today's post.
So the first thing to do to win the Canadian federal election is to register a Canadian political party. Shout out to anyone who registers the "Canadians hacking the election party" or "Canadians legally violating privacy party" to make even more obvious and tongue-in-cheek. Once you have a political party, you are ready to start legally collecting Canadians personal information.
As another aside, even if you don't go any further in terms of the steps below, you can now collect, use, and disclose any information on Canadians you want, and the federal privacy commissioner can't do anything about it as you're out of his jurisdiction! You could have access to a database of Canadian's mental and sexual health, and be selling it on the black market, and as a political party, I don't see anything that could be done to you. It should be obvious at this stage that the federal privacy commissioner should have jurisdiction to investigate any/all political parties, and on top of that, the ability to administer strict and real penalties for privacy violations. Back to this year's election:
So now you have your privacy law exempt party, and you need to collect some data. The easiest way to do this is to buy it, so you'll want some money. Think of any government, company, or person who might want full influence with the next political party in power -- reach out to them. There are a lot, this should not be challenging with a little creativity. I would say 10mm is an easy low number (in terms of taking control of a country like Canada), but obviously the higher the better. It should be no problem if you send them this article, and the laws haven't changed by then.
So now you've got a privacy law exempt political party, 10mm, and are ready to start collecting data. You have to remember, there's no one that can investigate you today in terms of privacy law, which governs the collection, use, and disclosure of Canadian's personal information. This means you can ask anyone who has access to any database of Canadians, the bigger the better, and you could offer to pay them for a copy.
In the last 5 years, I've had a data management head of one of the top 3 political parties tell me he was "given" a health related database. I asked him what he would do with, and he shrugged, "merge it with the rest of our datasets and see how it can help micro-targeting". This should demonstrate every step in this article has been used by someone, in some capacity, and I expect that each of these steps to be fully utilized over the next few months.
We're headed into the election cycle shortly, so you don't have much time here, collect all of the data sets you can. Merge them. Sanitize them. Sort them by address.
Now is where you prepare your social media strategy. This is where the story of Christopher Wylie and Cambridge Analytica comes in. What they needed to be able to accomplish their goals was an understanding of what the different target demographics of a voter look like.
I don't know the full details, other than speaking to Chris once or twice on this issue, aside from seeing him speak, and what I've read online, so this is a mixture of what I recall was done, and what I think should be done.
You launch some machine learning (what most people incorrectly call AI) chat bots, to start to create, and infiltrate as many discussions as you can. A basic example would be to start a conversation on anything that raises emotions in Canada, such as a gun registry, abortion, or immigration. It's pretty easy to guess which side of the political spectrum you're on based on that. In fact, even if you skipped by the "buy all of the data" step above, collecting data this way would still be quite effective. The power you will have from using machine learning on this data is you can start to see unexpected topic areas based on region. For example, the BC NDP endorse subsidies for LNG/Fracking, where as the AB NDP endorse oil/gas/bitumen/pipelines. That's at the provincial level. Once your machine learning discovers this, you would prepare different ads and messaging for NDP supporters in each province as a result. "Big data" allows you take this macro concept and break it down into micro-targeting.
Micro-targeting means instead of provincial level political beliefs, you could know the hot topic areas of your street, your apartment building, or even amongst a certain friends or family group. This data collection over the next few months is invaluable in two contexts. One is to score every voter, which requires you to know how political parties work. The other, is to now create communities.
Whenever a politician canvasses your door, unless they're new, they're not there to chat, they only want to know how likely it is you'll vote for them. For simplicity they'll give you a score from 1-5 after speaking with you, but each party does is differently. 1 would be that you'd never ever vote for that politician, 2 is likely not, 3 is unknown, 4 is a maybe, and 5 is absolute yes. Why this is done, is so that on election day, each political party starts with their 5s, and makes sure they get to the polls, even offering rides, once that list is gone through, then they go down through the 4s, etc. Until this generation, all of that was done at the door, but with the ability to access application programming interfaces (APIs), as well as website scraping, tools can be written to collect this data at scale. If you don't have someone knocking at your door this election season, you've likely already been scored.
The other thing that can be done with this data, and if I recall correctly, this is what Mr. Wylie and team did that was novel. They started to create real communities from digital ones. For example, if I saw that all the people with my political leaning in my neighbourhood were passionate about opiate addiction (We are! 43 overdoses last Friday alone, no joke!) and gentrification, I could start or join an online social group and start talking about these different issues. Now, I wouldn't make it obvious it's for a political party, instead it would appear just to be a group of like-minded people who think the same way I do about the same topics. This strategy goes even deeper, and this is what I've never heard being done before Mr. Wylie. At some point, he and his team were able to move these digitally created communities of like-minded people, many who didn't even realize this community was created for political alignment, and he would arrange a physical meetup in the real world. This would have to happen after a certain scale to not be detected, that the event was created by a robot. Let's say once a community of +1000 people existed, using a pseudonym, he could create an event at the local coffee-shop and say "Let's talk about these issues that matter in this neighbourhood!", and no one would realize that the event organizer might not have even been there as the event and group were digitally created!
Once you get people in a room talking about passionate topics, in the heat of an election cycle, you've activated a machine -- or in this case, +100s of groups of people.
At that point, you have your data team actively engaged on the digital side of these chat groups, not just in collecting all of the data, but dynamically steering the narrative towards the hot topics at that part of the election cycle that resonate with your voting demographic.
It's also worth noting which mediums are used by different demographics. For example, if you're looking at +40 year old right leaning Christian Conservatives, that's a demographic that's likely still more accessible on traditional television, as opposed to the 'gram.
The ability to shape, and more importantly to create, communities in the digital world is what Mr. Wylie credits to getting Trump elected, and Brexit/Leave.eu's success.
With online advertising, you can now micro-target. This means you could create a Facebook or Google ad that says "We promise electoral reform" which would resonate with me, but you could target women 18-27 in my building with an ad linking to the woman's right section of their political platform, and a different ad to the people on the top floor talking about airplane noise policy. Being able to micro-target to sway opinion will be huge in 2019, for those with the resources to do so.
Even if you don't follow these instructions, after reading this you will start to see ads that are micro-targeting you, so be aware of them. Anything your friends or neighbours think about SNC Lavalin today, will likely affect the ads you see tomorrow, as we tend to live in echo chambers digitally, and in the real world, sadly. This is one of the challenges in this politically polarized climate, the more passionate you are to the right or the left, makes you more vulnerable to micro-targeting.
That's about it! On election day, you'll have swayed opinions, created new communities, and you should have almost every voting Canadian in a database, with a rating from 1-5. On the week before election day, create an ad asking the 4s and 5s if they want a ride to their nearest polling station, and work your day down to the 3s and even 2s after that. Your goal is to get as many of your strong voters to the polls as possible.
Good luck, and remember me and this advice when you're in power for the next 4 years!
This may sound far fetched, but I've already been told I could make 6 figures over the next 6 months if I help a certain top 3 political party with their campaign. Every data expert with real capabilities you know, is likely to be asked to help.
I have several recommendations on how the government can minimize or stop this from being possible, as well as how to minimize the impact of your being manipulated as a voter if you want to reach out, otherwise I'll save those thoughts for another article.For further reading:
For further reading: