‘Anonymized’ Data Is Meaningless Bullshit

When most of us think of how the concept of “ data ” has been skewered by the press , we ’re plausibly conceive aboutan app ’s position datum tipping offour house address , orapps like Grindr tipping advertisers offabout our sexuality . What ’s less scrutinized , both by the publicand by those in public part , is data that ’s “ anonymized”—tied to something like an IP address , rather than a name — even though that ’s a concept we ’ve seen to be bullshittime and again .

The up-to-the-minute proof comes courtesy of Dasha Metropolitansky and Kian Attari , two Harvard studentswho recently built a toolthat combs through troves of consumer datasets upload from breaches across the web . As Metropolitansky and Attaritold Motherboard , their program was created to link together not - so - anonymous information — like emails or usernames — back to any “ anon. ” data that was found in a decade ’s worth of data point breach fromnearly a thousand different domains , from Adobe to YouPorn .

And — surprisal , surprisal — despite the mass of these datasets being “ anonymized , ” name someone caught up in a given leak was n’t hard at all , according to the researchers .

Photo: (Getty)

First , have ’s get some fact out of the way . Big shadowed datum factor , by and large , are n’t last to store anything explicitly personal about you — the individual reading this story — just because there ’s no note value in it . Even thoughthe advertizement stalk us around the webmight seem to suggest otherwise , trafficker give no hoot about your hopes , your dreams , your fears , the gym you go to or how you sexually identify — at least not on an individual grade . What they do wish about is catering a specific ad to a specific demographic , which is something that ’s ultimately harvest fromwhere you know , where you browse , and — yes , in some cases — whether you ’re queer - identified .

Here ’s a personal example : Based on my NYC - based paper trail — which involve purchases at Petco , Goodwill , andsome of my city ’s many gay bars , trafficker can realistically market me anything connect to khat , thrift shop , or anything bisexual with the confidence that they ’re not wasting money when aim me with ads . They do n’t need to bang who I am , per se — they just necessitate a way to reach the target demographic that I just so happen to be a part of .

Major data point agent havereams of aggregated intelon me that ’s fantastically worthful because it can plunk me into one of those demos with a surprising degree of truth . Any of these data points are n’t of necessity going to be tied to me , Shoshana , because they do n’t have to be to make other multitude money . What this data point is tied to might be something like my estimator ’s unequalled IP address ormy telephone ’s mobile advertizement identifier , which are , on their own , anon. .

Argentina’s President Javier Milei (left) and Robert F. Kennedy Jr., holding a chainsaw in a photo posted to Kennedy’s X account on May 27. 2025.

But even that particular information point is n’t really deserving that much — advertisers , on a day - to - twenty-four hour period groundwork , are look at my data ( and yours ) as it ’s aggregate with data from an untold bit of other masses . A person ’s single “ datum , ” on its own , ispretty much ugly ; after all , marketers ca n’t guarantee that I ’ll be clicking on a given advertizing or buying the product they ’re selling . What is valuable is when that data point ’s in aggregate , even if it ’s “ anonymized ” and not tied to any one somebody . This is why Facebook , for lesson , can say thatit ’s bring in roughly $ 26 a popfrom every user plugged into its system of rules — the only reason it can say that is because it ’s monitoring whatbillions of peoplein aggregate are doingon its platform and off .

While one data broker might only be able to tie my shopping behaviour to something like my IP computer address , and another broker might only be able-bodied to wed it to my scratchy geolocation , that ’s at last not much of an military issue . What is an issue is what happens when those “ anonymized ” data points inevitably bleed out of the marketing ecosystem and someone even more nefarious use it for , well , whatever — use your imagination . In other words , whenone data agent springs a leak , it ’s bad enough — but when dozens spring leaks over time , someone can nibble that data together in a fashion that ’s not only identifiable butchillingly accurate .

That ’s why the “ anonymized data ” defense from trafficker and data brokers is so fucked . It ’s a go - to line that they can technically twist to metre and again with a clean conscience , have intercourse that their own data appeal is by the books . At the same time , these are some of the same companies that have leakednearly 8 billion recordsover the past year , which in the end negates that logic in the first place . It ’s enough to make you wonder where the hired man - washing stops and the hand - wringing begins .

William Duplessie