Changing the Identifier May Not Change the Pattern

Updated: Oct 8
OBSERVATION 001
THE ASSUMPTION: Change the identifier and you break the link.
For a long time, identity was the convenient anchor. If two records shared the same subscriber identifier, account, device ID, or address, linking them was straightforward. So a lot of privacy and misattribution engineering has focused, reasonably, on removing, rotating, or obscuring those anchors.
That still matters. But it solves a narrower problem than it used to.
Correlation no longer has to begin with a durable name tag. Two observations can look unrelated at the identifier layer and still look remarkably similar everywhere else. Geography persists. Movement has continuity. Devices have characteristics. Network relationships repeat. Applications behave in recognizable ways. Events happen in sequence. Enough weak signals, observed together, can rebuild confidence that two different records describe the same thing.
This changes the question. Instead of asking, “Can the observer see the same identifier twice?” we should also ask, “What would make the observer believe these two observations belong together?”
That is a harder test because it is probabilistic. There may be no single smoking gun. The pattern itself becomes the evidence.
The implication is not that identifier protection has stopped working. It is that its success should be measured against the environment it operates in. Breaking one linkage mechanism is valuable. Breaking the observable continuity around it is a different problem.
Misattribution is moving from identity management toward inference management. The identifier is still part of the picture. It just may not be the whole picture anymore.
WHAT WE ARE WATCHING
Whether privacy and misattribution architectures are being tested against correlation across layers, not only against recovery of a protected identifier.




