When the Database Forgets to Forget: The Curious Case of Right-to-Erasure Failures
We often talk about digital preservation as a heroic act: a fight against the slow decay of bits, a race to save precious cultural artifacts from the abyss of forgotten servers. But what about the opposite problem? What happens when we *want* something to be deleted, and the mechanisms of the web, designed for permanence, refuse to let it go? This is the curious and often troubling reality of right-to-erasure failures, a paradox at the intersection of privacy law and digital preservation.
Laws like Europe’s GDPR established a powerful principle: the ‘right to be forgotten.’ It suggests that individuals should have some control over their digital footprints, able to petition for the removal of personal data that is no longer necessary or relevant. On the surface, this is a victory for individual autonomy. The reality, however, is a murky battleground where deletion is rarely absolute. The very features that make the web a robust archive—redundancy, caching, and public indexing—actively work against the goal of complete erasure.
Consider a simple example: a local newspaper archives an article from twenty years ago that mentions a person in a minor, perhaps embarrassing, context. That person, seeking a fresh start, successfully exercises their right to have the article removed from the newspaper’s primary website. A clear victory. But the story doesn’t end there. The Internet Archive’s Wayback Machine, operating as a public service, may have captured that page days, weeks, or years ago. Aggregator sites may have scraped the headline. A PDF of the article might exist on a university server, saved by a researcher studying local history. The data has proliferated, spawning copies in independently managed silos far beyond the reach of the original publisher.
This creates a fundamental conflict of values. The individual’s right to privacy clashes with the archivist’s mandate to preserve a complete and unaltered historical record. Which is more sacred: a person’s past or the past itself? There is no easy answer. Archiving advocates argue that altering the historical record, even for laudable reasons, sets a dangerous precedent and creates a distorted view of history. Privacy advocates counter that forcing individuals to be eternally haunted by digital ghosts is a form of punishment without end.
The problem is not that the systems of preservation are malicious; it’s that they are efficient. They were built to remember, not to forget. A web crawler doesn’t ask for context or intent. It sees a public link and saves it. This technological ‘absent-mindedness’ means that the promise of the right to be forgotten is, in many cases, a legal fiction. True deletion requires a level of coordination and authority that simply doesn’t exist across the global, decentralized web. We are left with a landscape where data can be both deleted and preserved simultaneously, a Schrodinger’s cat of personal information, leaving us to wonder if anything on the web can ever truly be forgotten, even when it should be.
Notes & further reading
A few pages I came back to while writing this: