Facebook stands alone in the breadth and depth of personal data it collects.

Since Donald Trump's election, many in the tech industry have been concerned about the way their skills—and the data collected by their employers—might be used. On a number of occasions, Trump has expressed the desire to perform mass deportations and end any and all Muslim immigration. He has also said that it would be "good management" to create a database of Muslims, and that there should be "a lot of systems" to track Muslims within the US.

In the final days of his presidency, Barack Obama has scrapped the George W. Bush-era regulations that created a registry of male Muslim foreigners entering the US—the registry itself was suspended in 2011—but given Trump's views, demands to create a domestic registry are still a possibility.

As a result, some 2,600 tech workers (and counting) have pledged both not to participate in any such programs and to encourage their employers to minimize any sensitive data they collect. The goal is to reduce the chance that such data might be used in harmful ways.

The fear in the tech community is of being complicit in some great crime. The neveragain.tech pledge reads, in part:

We have educated ourselves on the history of threats like these, and on the roles that technology and technologists played in carrying them out. We see how IBM collaborated to digitize and streamline the Holocaust, contributing to the deaths of six million Jews and millions of others. We recall the internment of Japanese Americans during the Second World War. We recognize that mass deportations precipitated the very atrocity the word genocide was created to describe: the murder of 1.5 million Armenians in Turkey. We acknowledge that genocides are not merely a relic of the distant past—among others, Tutsi Rwandans and Bosnian Muslims have been victims in our lifetimes.
Today we stand together to say: not on our watch, and never again.

Their concerns are not unfounded. IBM, in particular, has a dark history when it comes to assisting with genocides. The company's punch card-based Hollerith machines were instrumental in enabling the Nazis to efficiently round up Jews, seize their assets, deport them to concentration camps, and then systematically slaughter them.

After Trump's election, IBM CEO Ginni Rometty wrote the president-elect to congratulate him on his victory and offer IBM's services in support of his agenda. Oracle co-CEO Safra Catz has joined Trump's transition team, rank and file workers have been outspoken in their unwillingness to cooperate with programs that don't, in their view, respect the Constitution or human rights or which have disturbing historical precedent. Rometty's letter has provoked a petition from current and former IBM staff; Catz's role has resulted in at least one resignation.

One company, however, stands head and shoulders above the rest when it comes to collecting personal data: Facebook. Facebook's business is data collection in order to sell more effectively targeted advertisements. While massive data collection is not new or unique to Facebook—search engines such as Google and Microsoft's Bing have the same feature—Facebook is unusual in that it actively strives to make that information personally identifiable. Facebook accounts tend to use our legal names, and Facebook relationships tend to reflect our real-life associations, giving the company's data a depth and breadth that Google or Microsoft can only dream about.

Among the pieces of personal information that the site asks users for is religion. As with most pieces of information that Facebook requests, this is of course optional. But it's an option that many people fill in to ensure that our profiles better reflect who we are.

This data collection means that Facebook already represents, among other things, a de facto—if partial—Muslim registry. Facebook has the data already; the company can provide a list of self-attested Muslims in the US simply by writing a query or two. That data could be similarly queried for anyone who isn't straight.

As such, government coercion of Facebook—or even a hack of the company—represents a particular threat to civil liberties. Accordingly, Facebook should take a simple and straightforward protective step: delete that information.

Deleting the information will not make Facebook safe. It will still be a treasure trove of relationships and associations, and an intelligence agency could make all manner of inferences from the data contained within. (Religion, for instance, is likely to be discernible from the content of posts and from images of holidays and religious gatherings, but this would be more difficult to do in bulk—though we know similar inferences are already made about race.) But it would mean that Facebook is no longer so trivially searchable, and it would mean that it ceases to be such a clear database of religious affiliation.

Making a change like this should be trivial for Facebook. No doubt it would marginally reduce the company's ability to tailor advertisements to individual users—but it would serve as a clear statement against the threat such a database poses.