Our Data Privacy and the Issue With Inferences

How much would “owning” your data actually protect your privacy?

Host Kirsten Martin is joined by Ignacio Cofone, an assistant professor and Canada Research Chair in Artificial Intelligence Law & Data Governance at McGill University’s Faculty of Law. His research focuses on privacy harms and on algorithmic decision-making, with his current projects examining how to evaluate standing and compensation in privacy class actions and how to prevent algorithmic discrimination.

Ignacio came on the show to talk about his paper “Privacy Standing,” which appeared in the University of Illinois Law Review.

Providing courts with guidance on how to assess privacy injuries and advocating for people’s rights to seek compensation for them (i.e., legal standing), Ignacio’s paper distinguishes between what constitutes a privacy loss, a privacy harm, and an actionable privacy injury. He also seeks to define downstream, consequential harms as something distinct from privacy harms so that the latter can be recognized as harmful on their own and not dismissed simply because they haven’t (yet) led to something more tangible like identity theft or a financial loss.

As for where privacy harms originate, Ignacio emphasizes how frequently they arise not from the moment our data is collected but rather from the inferences later made about us from that data—or even from the data of others who just happen to be similar to us. That means the prevalent approach of giving people notice and choice—which Kirsten traces back to the economics of information literature of the 1960s—and its focus on asking users for permission to collect their data is in many ways inadequate when it comes to protecting our privacy.