In an episode of the speculative-fiction TV series Black Mirror, after a woman’s boyfriend dies in a car accident she discovers a chatbot that can use her dead partner’s “digital remains” to enable her to simulate talking with him. Because, like most people in the modern world, he created thousands upon thousands of emails, social media posts and the like to sift through, the bot could recreate not only his voice and tone but also his sensibility, his humor, his beliefs — almost his whole personality. The woman’s initial fear of the technology gives way to longing, which makes her ripe for a “premium” version that inserts these capabilities into a lifelike replica of her boyfriend who will sit with her at breakfast, walk with her in the garden and, yes, have sex with her.

Several companies — among them, Silicon Intelligence — have made elements of this story a reality by creating “digital ghosts” or “griefbots” that virtually resurrect the dead. Using AI, explains Katarzyna Nowaczyk-Basińska, a University of Cambridge researcher, technologies can “simulate language patterns and personality traits based on processing huge amounts of personal data.”

The Black Mirror plot of technologically cheating mortality hints at myriad issues concerning what Swedish political scientist Carl Őhman calls “the afterlife of data” in his recent book of that name. As we fire off scores of emails and messages every day, venture onto Instagram or TikTok, bank or date online, and shop on Amazon, rarely do we think about what will happen to the records after we die. Yet, barring massive cyberwarfare, the demise of Alphabet and Meta (the parent companies of Google and Facebook), the nuking of data centers, or the conscious decisions of our heirs, corporations or governments, our information will live on.

A whole lot of it, too. Humanity produces 147 zettabytes (147 followed by 21 zeroes) of data per year — in Őhman’s understatement, the “the largest archives ever assembled.” And some 167,000 data-creators (i.e., people) die every day — another two billion or so by mid-century.

Data, Data Everywhere …

The immortality of data, or at least its daunting longevity, raises a host of issues ranging from ethical to economic to legal. And these are largely ones that are not discussed by economists or business leaders, philosophers or politicians. Or anyone.

To begin, consider what becomes of personal data when someone leaves this mortal coil. While digital photos, like old-style photo albums, might reasonably be things that a child, relative or friend might inherit, can posts on Facebook or Reddit, video-game avatars or swipes on Tinder be heritable? Are such data things — or something less tangible, morsels of identity? Is it the property of family and friends or of the corporate behemoths like Apple or Amazon that archive them? Or does it fall into a murky sort of public domain, to be mined by marketers, governments, historians, journalists or run-of-the-mill snoops?

This much is clear: the dead have no legal right to privacy, so embarrassing (or worse) secrets that once would have gone to the grave can posthumously harm reputations or torment descendants. Because people often use search engines as a sort of confessional or a playing field for the id, unarticulated worries and unacted upon desires now leave a vast trace of zeroes and ones. Who would not be embarrassed by what Uncle Elmer left in his deleted email folder? 

Managing Your Digital Legacy

In theory, before they die individuals have the option of wiping out their data, or at least any materials they don’t want the world to see. But who does this, or even knows how to do this? And even if one eradicated online accounts, most people (correctly) suspect that they survive somehow, somewhere in zombie form. Another option is to use online platforms like Wills.com, which advertises that it provides “detailed instructions for your digital executor” so that “your digital legacy is managed according to your wishes and provides peace of mind for your loved ones.” Some estate lawyers also offer digital wills, enumerating what digital information is accessible to whom. 

Heirs could try to take down online accounts and erase other personal materials, provided they know their antecedents’ array of accounts and passwords. But even if they did know enough, it’s possible, if not likely, they would happen upon something that could hurt or shock them.

In any case, the afterlife of data may affect the living in other ways. How many times have you gotten notices from Facebook to wish a dead “friend” happy birthday, or been treated by Apple to a cascade of jolly photo memories of an estranged loved one? Őhman speculates that the persistence of digital detritus may prolong grieving or delay “closure” to a terminated relationship.

Of course, online data has the potential to preserve and enrich good memories as well as deepen understanding of the deceased. A favorite song on Spotify or even the receipt of a purchase of a birthday gift could serve as digital madeleines. Proust lives … sort of. 

There are wheels here within wheels. Islamic prayer apps such as Du3a.org enable subscribers to set up prayer supplications that will regularly go out to their descendants after they die. SafeBeyond, one of several startups in the “digital afterlife industry,” offers “posthumous delivery of digital, video, audio and written messages to pre-selected recipients, in accordance with the specific and predefined personal requirements of the user.” A long-dead grandparent could speak at their granddaughter’s wedding — or a deceased neighbor could email daily complaints about a barking dog.

While these companies have yet to make a splash, Facebook already offers a “legacy contact” feature and Google an “inactive account manager.” Őhman rightly worries, however, that profit-driven businesses may be insensitive or “unreliable guardians of digital heritage.” It certainly seems a stretch to imagine that data managers and programmers will apply the Golden Rule to their treatment of iffy data.

Őhman wonders whether it would be better to assign such responsibilities to governments or NGOs. But then he rightly pivots to George Orwell and the Ministry of Truth he wrote about in 1984: “‘Who controls the past,’ ran the Party slogan, ‘controls the future.’”

Given these dangers, should there be laws or regulations protecting what British researcher Edina Harbinja calls “postmortal privacy”? The EU’s landmark General Data Protection Regulation, enacted in 2018 and copied from Japan to California, safeguards individuals’ digital privacy and enshrines the “right to be forgotten.” A few countries, notably France, have issued regulations broadly pronouncing that an individual can “define general or particular guidelines regarding the retention, deletion and communication of his/her personal data after death.” However, neither the U.S. nor the EU have clear statutory or case laws governing postmortal privacy.

• • •

Consider, while we’re adding up the pluses and minuses, that the accumulation of digital remains could also be a boon to historians, policymakers, and society. As the late George Mason University scholar Roy Rosenzweig predicted in 2001, Internet-based data will offer researchers an “essentially complete historical record.” 

If one believes that learning from the past improves our chances for a better future, such knowledge could prevent wars, mitigate pandemics and reduce poverty. Indeed, riffing off this notion, Őhman wonders how we might be illuminated if we had Churchill’s online search history, Napoleon’s tweets or Jesus’ YouTube clips.