By Caroline Santoro
How could the campaign use Facebook data to predict political preferences? If you saw my Journal Club presentation on February 22nd, you know that the machine learning algorithm referenced in the paper “Computer-based personality judgments are more accurate than those made by humans” could predict scores on a personality test better than a person’s friends and could predict political preferences equally well. To make the algorithm, the researchers created a Facebook app called “myPersonality,” in which they gave participants a standardized personality test in order to assess the “Big Five” Personality Traits. The researchers ran linear regression models and trained their machine learning algorithm to predict not only scores on the test, but also “external validity” items like political preferences, substance use, and physical health (Youyou, Kosinski, Stillwell, 2014). Just as this study used Facebook “likes” to predict political preferences, Kogan could predict the political preferences of the 50 million people on whom he had data. In turn, the Trump campaign could use these preferences to better tailor political messages for these people.
This digital disaster raises several compelling ethical issues. How was Cambridge Analytica able to use for political purposes data that was collected exclusively for academic use? Why was the data not anonymized? The people could be targeted only because the data included profile and contact information. Moreover, if only 270,000 participants gave access to their data, how did 50 million more people have their personal data taken without their consent? Facebook stresses the importance of privacy, and frequently requests that users review their privacy preferences to limit who can see their information and posts. Do none of those limits matter if someone is willing to pay for the data?
Mark Zuckerberg, Facebook founder and CEO, wrote a post on his Facebook page on March 21 addressing the “Cambridge Analytica situation.” He outlined a timeline of events and detailed what Facebook had done and what they are doing to prevent this from happening again. He also took responsibility for the data breach, declaring, “We have a responsibility to protect your data, and if we can’t then we don’t deserve to serve you…I started Facebook, and at the end of the day I’m responsible for what happens on our platform,” (Zuckerberg, 2018).
Digital scandals like these, alongside fears of the growing capabilities of machine learning and artificial intelligence, raise the need for a “digital ombudsman,” an oversight organization that could ensure that companies respect privacy terms and ethically use technology and data. Applications like Facebook rely on users’ trust that personal data will remain out of the hands of those who would exploit or harm them. After the scandal with Cambridge Analytica, what will Facebook do to regain that trust?
Chang, A. (2018, March 23). The Facebook and Cambridge Analytica scandal, explained with a
simple diagram. Retrieved March 24, 2018, from
Murray, S., Reston, M., Bash, D., & Perez, E. (2018, March 21). Inside the Trump campaign’s
ties with Cambridge Analytica. Retrieved March 24, 2018, from
Mark Zuckerberg. (2018, March 21). Retrieved March 24, 2018, from
The Economist. (2018, March 22). The Facebook scandal could change politics as well as the
internet. The Economist.