A Cambridge University academic obtained data from Facebook users with a widely used tool provided by Facebook. He gave us assurances that this dataset could be lawfully licensed to be used for marketing purposes.
Dr Aleksandr Kogan of Cambridge University undertook a large-scale Facebook research project on Cambridge Analytica’s behalf in the United States in 2014.
Cambridge Analytica decided to seek the expertise of Aleksandr Kogan because he was an academic at an internationally renowned university who routinely undertook these kinds of research projects. He had published articles on the subject of personality profiling in academic journals, and had already created a data sampling methodology. Cambridge Analytica had no experience in developing Facebook apps.
His company, Global Science Research (GSR), had built an app that collected data from Facebook profiles with a tool that was provided by Facebook and was commonly used by app developers at the time. It allowed for the collection of information on both respondents and their friends.
At the time, Facebook’s terms of service for users stated that when a user granted access to such an app, “your agreement with that application will control how the application can use, store and transfer that content and information.” GSR and Dr Kogan’s terms of service requested the consent of users for the transfer or sale of their data.
Dr Kogan made contractual commitments to Cambridge Analytica that GSR had acquired this dataset in accordance with all applicable laws, including data protection legislation.
Our contract with GSR/Kogan was only for users in the United States, and not for users in any other countries. Facebook has reported that up to 87 million may have had their data affected by GSR’s work. We licensed data for no more than 30 million people, as clearly stated in our contract. We received no more data than this.
The data consisted of public Facebook profiles and scores (mostly estimated) on personality traits produced by GSR. The research company did not share the content of any private messages with us, and we do not handle such data.
Tests showed that the Facebook data collected by GSR was not helpful for marketing campaigns, and Cambridge Analytica did not use the data on the Trump campaign.
Cambridge Analytica provides a range of marketing services to brands and political campaigns. We tested GSR data in 2014 and 2015 to see if it was possible to define categories of people with similar personalities in a way that could improve the performance of advertising campaigns.
When we studied the effectiveness of advertising using the GSR personality types, as compared to more conventionally targeted advertising, by comparing the number of clicks for every dollar spent, the GSR data was shown to be less effective than standard off-the-shelf demographic segments.
We decided in 2015 to collect our own personality data through market research panels, with the full consent of each respondent, so that we had more accurate data with which to continue our work.
Due to time and budget considerations, instead of using such personality-based targeting on the Donald Trump campaign in 2016, Cambridge Analytica provided more conventional data modeling for donors, persuasion and turnout.
We did not share the GSR data with any other party.
We deleted the data after Facebook told us in 2015 that GSR had broken their terms of service. No traces exist of it in our work.
When Facebook told us in 2015 that GSR had broken their terms of service, we deleted the raw data from our file server and began the process of searching for and removing its derivatives in our system, in coordination with Facebook. No traces exist of it in our work.
Subsequently, we took legal action against GSR for breach of contract.
We cooperated with Facebook to ensure they were satisfied that we had not knowingly breached any of their terms of service. When Facebook sought further assurances in 2017, we carried out an internal audit to make sure that all the data, all its derivatives, and all backups had been deleted.
Because of recent inaccurate news reporting, we will be undertaking an independent audit by a leading data forensics firm to demonstrate that we do not hold any GSR data.
We have been working with the UK’s Information Commissioner’s Office (ICO) since February 2017, when we hosted its team in our London office to provide total transparency on the data we hold, how we process it, and the legal basis for us processing it. We offered to share all communications and documents we hold relating to GSR with the ICO, and to help them search our systems for GSR data.
The ICO said in May 2017 that it would "open a formal investigation into the use of data analytics for political purposes", and engage “with a range of organisations – political parties and campaigns, data companies and social media platforms".
The ICO’s remit is compliance with existing laws, identifying any limitations in existing legislation, and generally raising public awareness and confidence: these are all goals which Cambridge Analytica supports.
We did not use the GSR data for the Trump campaign. We used RNC data, providing the campaign with the same kind of political data program as used by the Obama and Clinton campaigns.
Cambridge Analytica provided polling, data analytics and digital marketing to the Donald J. Trump campaign as part of a collaborative effort under Brad Parscale, alongside the RNC and other vendors. The company was hired by the campaign in June 2016.
After initially using our own commercial and political data as well as our own suite of models, our data team used the following sources of data from August onwards:
We used the data to identify “persuadable” voters, how likely they were to vote, the issues they cared about, and who was most likely to donate. We also built a polling tracker for every key state, and provided dashboards for the campaign, including the group that planned the candidate’s travel agenda.
We used the same kind of political preference models as Obama and Clinton campaigns; however we started five months out from election day with far fewer resources and less data. It’s rare to have one vendor working across so many different campaign functions – we integrated polling, data science and marketing into a single operation. Having a large amount of control over each of these three areas allowed us to be extremely efficient and reactive.
Our analysis was also used for targeted advertising. We managed a large proportion of the digital advertising budget on behalf of Giles-Parscale in the context of the Trump campaign. In doing this we used a suite of models produced by the data science team, which outlined profiles such as undecided voters or inactive supporters, and matched these audiences to online cookies, mobile devices, and social IDs. Onboarding data through ad tech platforms is standard digital advertising practice.
We also relied on the audience segments that Facebook and other online platforms make available to all advertisers. These are based on interests and demographics, to help serve the most relevant ads to the most relevant people. Our digital marketing used core campaign messages with "paid for by" disclaimers to persuade voters to vote, increase turnout among supporters, and boost volunteer numbers and donations. In the case of Facebook and Twitter ads, these were clearly linked to the official Trump presidential campaign Facebook and Twitter accounts. We did not use bots.
We didn’t have the opportunity to get into personality models. Building a presidential data program often takes campaigns well over a year. We focused on the core elements of a political data science program, as we explained following the election at public events, in media interviews, and on our website.
Elections are won or lost by candidates, not data science. Data is important in modern campaigns for deciding how to allocate resources and for making advertising more efficient, but of course the candidate and their message ultimately needs to connect with the electorate.
We held preliminary meetings with Leave.EU to explore working together, but we delivered no work and the project did not go ahead.
It’s a matter of public record that Cambridge Analytica held discussions with Leave.EU and submitted a business proposal to Leave.EU to help them secure the designation as the official Leave campaign. Leave.EU were very keen to talk publicly about an association with us. They were vying for nomination as the official Leave campaign; working with a data analytics firm with experience in US presidential primaries would have given them credibility. We agreed to share a platform with their representatives to talk about the kind of work we could deliver in the future.
We did not move forward with this project. No contract was signed, no payment was made, and no work was delivered to any party or campaign group.
Ultimately we did no work on any referendum campaign. The suggestion that we were somehow involved in any work done by Aggregate IQ (AIQ) for Vote Leave in the 2016 EU referendum is entirely false.
For clients completely unrelated to Leave.EU, we subcontracted some digital marketing and software development to AIQ. The UK parliament’s Digital, Culture, Media and Sport Committee has published documents, including intellectual property licences, which show that Cambridge Analytica hired AIQ in 2014 and 2015 to build campaign management software.
The Guardian newspaper has printed a retraction following its allegation that AIQ was a direct part of Cambridge Analytica/SCL, and that AIQ and Cambridge Analytica secretly coordinated on the EU referendum.
The media storm originates from claims made by a former part-time contractor who has completely misrepresented himself and the company to the press. Christopher Wylie is not a whistleblower.
Mr Wylie has been the source of much of the media’s reporting on Cambridge Analytica. Much of his account is based on conjecture and guesswork – he admits himself that what he says is speculation – while his own motivations in this saga have remained unexplored.
Mr Wylie has repeatedly claimed to be a founder of Cambridge Analytica. In reality, he was a part-time contractor for SCL Elections for a period of 11 months and left in July 2014. He has no recent knowledge of our business or its practices, and has admitted as much in his testimony to the UK parliament.
He set up a rival company called Eunoia Technologies, which pitched micro-targeting data science services to the Trump campaign and later tried to work for Vote Leave. Our lawyers took action against Mr Wylie to prevent his misuse of the company's intellectual property while attempting to set up his own rival firm. He has now rebranded himself as a “whistleblower” following his failure in the industry.
We believe that we should all have more control over our data, and there should be more transparency over how and when it is used.
We use data fairly and transparently:
Our data is secure:
We believe that we should all have more control over our data, and there should be more transparency over how and when it is used. We welcome Europe's new data protection laws (GDPR). There are very good reasons for updating current data regulations, which date back years to a very different time. From giving everyone more protection, to promoting a more equal privacy landscape, these changes will be good for the industry as a whole.
While Cambridge Analytica entered the US market by working for Republican clients, globally we work across the mainstream political spectrum.
Recent media coverage has sought to portray the company as somehow politically or ideologically driven. As anyone who is familiar with our staff and work can testify, this couldn’t be further from the truth.
Since its inception, Cambridge Analytica has grown into a data-science consultancy and marketing agency focusing mainly on the commercial market. The vast majority of our business is commercial rather than political, contrary to the way some sections of the media have portrayed us.
Our previous political work has covered both left-of-center and right-of-center candidates around the world.
We don’t engage in or condone hacking, entrapment, or so-called "honeytraps", and nor do we use untrue material for any purpose.
SCL Elections, the precursor to Cambridge Analytica’s global political division, has been accused of unethical practices. The Board of Cambridge Analytica has asked a senior British lawyer, Julian Malins QC, to lead an independent investigation into these allegations, the findings of which will be shared publicly in due course.
The Board also suspended CEO Alexander Nix with immediate effect, pending the outcome of the investigation. In the view of the Board, Mr Nix’s comments secretly recorded by Channel 4 News did not represent the values or operations of the firm and his suspension reflects the seriousness with which they view this violation.
We don’t engage in or condone hacking, entrapment, or so-called "honeytraps", and nor do we use untrue material for any purpose.
Like any marketing agency, Cambridge Analytica uses social media platforms for placing paid advertisements and organic content. Influencer marketing and building grassroots networks on social media are both common activities for a modern political campaign.