By Raven Jiang
Datafication of the economy has brought us many good things, from cheaper flights to two-day shipping on Amazon Prime, but it is also a great source of controversies. Beyond the obvious Orwellian concerns that we often hear in recent discussions over online privacy, government surveillance and Facebook social experiments, there are more subtle battles over the implicit balance of power. So, how do we draw the line between a company using customer data to improve their customers’ user experiences and a company using that same data to improve its own welfare, possibly to the detriment of the consumer. As it stands, there is an imbalance of power in the industry that allows the latter to pass off as the former.
When we give our data to Google, either intentionally by filling out forms or passively through using Google products, we are providing valuable information that Google can convert to advertising revenue. Unlike a delivery service using our data to improve delivery times for us, Google is using our data to improve a service it provides to third parties. The argument in support of doing so is that advertising is at the very least providing an indirect benefit to us (by paying for the free products we enjoy), if not a direct one (by showing us things we want to buy). There are some people who reject this and conscientiously avoid Google products, but the rest of us all desperately want to believe that the argument is valid because not using Google would be too hard.
Still, most people would agree that having their private emails read by Google would be violation of expected norms. Yet, Google has repeatedly argued in court that its automated mining of emails for targeted advertising is not illegal wiretapping because it is part of the “ordinary course of business.” Perhaps it should be perfectly legal then for Google to also run automated analysis on the private emails of job applicants to determine their fit to work at Google. If targeted advertising is in the consumer’s interest, a more informed interview process should really be a great thing for the potential employee. After all, someone who is rejected as a result of automated email screening ought to be glad that he/she did not mistakenly end up working for a company that is not the “right fit.” Of course, no one would actually buy that argument, because we do not want Google to have that kind of power over us. We draw the line somewhere when it comes to the insights a company is allowed to draw from our data.
LinkedIn is a company that has collected massive amount of data about individual employees. The interesting thing about LinkedIn is that it bases its internal decisions on business intelligence about other companies gained from analyzing its user database. For example, when selling to another company, LinkedIn’s sales people can figure out, based on knowledge about the company’s organizational structure, who in the company’s human resource department to target. LinkedIn also uses its own product in its recruitment effort. In other words, when tech workers willingly offer their personal data to LinkedIn, they are in some sense conferring power to LinkedIn and reducing their own employers’ capacities to compete with LinkedIn.
These are relatively benign examples, but they highlight the temptation that all tech companies have in using data not just to offer better services but to better themselves. Hypothetically, it should be possible for LinkedIn to figure out the future plans of other competing tech companies simply by looking at the skill sets of their recent hires. But is it ethical for LinkedIn to actually do so? LinkedIn profile information is ostensibly public, but only LinkedIn has direct access to the data, limiting the capabilities of third parties to draw the same conclusions. LinkedIn alone can determine the format and protocol through which our profiles are accessible, and being the gatekeeper of our data gives LinkedIn the power to create commercial value.
There are strong moral hazards at play when it comes to companies and the data we give them. For now, most companies are still focused on competing by creating better products and services with their data, but that will not last. In the great debate over government surveillance, we forget that for all the failings of the FISA court, there exist formal mechanisms in which government’s temptation to use data it has access to in unintended ways can be kept in check. If Google or LinkedIn ever decided to use their user data for more unambiguously selfish purposes, there seems to be little we can do about it. In the age of big data where every database is being turned into a commodity for sale, expectations of privacy and social norms have clearly failed to keep in check the power of data collectors.
Contact Raven Jiang at jcx ‘at’ stanford.edu.