Reduce exposure
Use tracker blocking and browser privacy controls where they fit the context. Remove continuity where continuity is not serving you.
The turn
The absurd part is not the meter.
The absurd part is that more systems already make this distinction quietly, constantly, without showing the instrument panel.
The omitted surface
This page shows a few signals because a page has edges.
Production systems do not stop there.
They combine context, device state, timing, navigation, repetition, history, failure, recovery, and third-party inference.
The visible cursor is only the first surface.
The claim is not that every site uses every technique. The claim is that the collection surface is broad, composable, and often invisible. This page is unusual only because it labels the apparatus.
The registry exists for a reason
California does not maintain a public registry for imaginary industries. AB 1202 created a data broker registration regime; the Delete Act, SB 362, expanded it and moved administration to the California Privacy Protection Agency. The registry and DROP do not describe the whole ecosystem. They make one point plain: personal data brokerage is ordinary enough to register, publicize, and route into common deletion infrastructure.
The names to examine are not obscure: Acxiom/LiveRamp, Experian, LexisNexis Risk Solutions, CoreLogic, Epsilon, people-search services, and location-data firms appear in adjacent markets depending on product line and legal definition. They are not all the same actor. They show that identity, risk, audience, and location data have ordinary corporate form.
Signals beyond the demo
What is omitted here is larger than what is shown.
Device and browser characteristics
Session continuity and cross-session consistency
Navigation rhythm and path repetition
Failure patterns, retries, login, and recovery behavior
Timing anomalies and automated regularity
Touch, mobile, or sensor-derived interaction
Environmental and network context
Prior risk history, linked identities, and vendor correlations
No single signal has to decide the case. Weak traces accumulate. Continuity is often inferred, not declared.
Where these signals go
Collection does not stay in one box. A usable signal tends to move.
- Analytics and replay platforms
- Flow repair queues
- Experimentation and funnel analysis
- Fraud detection and risk scoring
- Account recovery and challenge systems
- Abuse and moderation controls
- Vendor systems and third-party processors
Not all uses are malicious. Some are defensive. Some are operational. Some are extractive. Many are all three at once. The same trace can block fraud, raise conversion, and deepen extraction.
Evidence exhibit
This is what the signals become.
The exhibit shows the structure produced when public records, inferred attributes, purchased identifiers, session telemetry, and modeled scores are arranged into one operational profile.
It is not a side page. It is the evidence object this chapter has been pointing toward.
What people can do
Invisibility is not the condition. Control is.
Separate contexts
Separate identities, browsers, or profiles for work, finance, health, research, and ordinary browsing when the stakes differ.
Increase friction selectively
Grant persistent login, recovery, cross-device continuity, and background access only where the trade is explicit.
Read the purpose
Anti-fraud and accessibility collection are not the same as ad-tech extraction, even when the same signals are used.
Some systems need confidence. Not every system deserves intimacy.
Rights are uneven, but not absent. Where law provides a mechanism, use it; California now offers DROP as a one-request deletion route for registered data brokers.
Proof without discipline becomes another dossier.
Stolen facts are simple to replay. Presence is harder to fake. That makes it useful, but it also makes the collection more sensitive. Discipline is not optional.
The problem is not that presence exists. The problem is that it is often captured without discipline. What matters is not only whether a system can measure you, but whether it is allowed to keep, combine, sell, or repurpose what it learns.
A restrained proof system should collect less, reveal less, retain less, and still function.
Proof should be purpose-bound, proportionate, and resistant to replay. It should justify its scope, constrain its storage, and separate confidence from intimacy.