As people log in into their user profiles and share likes, emoticons and witty comments with friends and family, they might reveal some personal information to the social media between the lines. At first glance, all of that seemed harmless until Cambridge Analytica scandal brought to light the severity of social media fraud and sparked a privacy awakening.
Data is the new oil
Even though debates on obtaining and using data are not stopping in the meantime, it surely is social media’s fuel. Of course, the internet giants also gain advantage from data, like astronomical numbers of customers and sprawling cloud networks built to their own specifications.
In order to gather data and consequently become more “net-worthy”, social media is using AI assisted psychographic targeting.
What is psychographics? Psychographic studies examine and classify people according to their attitudes, aspirations focusing on understanding cognitive attributes, such as customer emotions, values, and attitudes, among other psychological factors. It involves asking specific groups of consumers questions about their activities, interests, preferences, and opinions. Researchers can then blend the results with other datasets—such as demographic or geographic information—in order to develop a more nuanced portrait of the user group (Huffpost).
Without asking questions and conducting surveys, data i.e. people’s internet activity is basically recorded by “cookies” which can be utilized to understand users in order to reach them more precisely and more effectively. Jerry Kaplan, Artificial Intelligence expert, discusses how companies can effectively determine (and monetize) your preferences and behavior by slicing and dicing this information into datasets.
The question is how far do psychometrics go?
Cambridge Analytica Scandal
Cambridge Analytica, the UK-based political consulting firm that was involved in Trump’s campaign, used harvested raw data exposed by Facebook from up to 87 million profiles.
The company’s data scientists used a simple online quiz (based on a personality test from the University of Cambridge) to flag respondents along dimensions of the Big Five Model i.e. the OCEAN model which tests openness, conscientiousness, extraversion, agreeableness, natural reaction or neuroticism, and predicted their reaction to political advertising.
This allowed Cambridge Analytica to reverse engineer a person’s personality profile using only their Facebook activity: pages and comments they liked, images and videos they chose to share and upload, etc. The accuracy of said profile was deemed extremely high when it comes to predicting an individual’s behaviour based on personality traits.
After the scandal, this is what the users have done:
Psychographics combatting payment fraud
Psychographic targeting is, however, not all bad. With this type of behavioural biometrics, banks and payment processors can use an individual’s unique pattern of behaviour while using their mobile devices like swiping and tapping to make sure the right person has made certain payments.
In that context, Mercury Processing Services International is now working on a more sophisticated machine learning model that will be regularly retrained and maintained internally.
Social media networks like Facebook are yet to be regulated because online data usage is still new and not included in regular legislations. Internet security and information theft is not to be taken lightly with the many scandals that happened in the recent past.
Luckily, new directives such as GDPR and PSD2 are helping a lot in regards to security in payments and also protecting users online.