Privacy for Sale

The unsettling premise behind Fog Reveal is not that police secretly installed a tracker in your pocket. It is that millions of people installed the trackers themselves—in the form of ordinary smartphone apps—and an advertising ecosystem transformed those location signals into a commercial product that could later be queried by government investigators.

The original source essay behind this article describes how Fog Data Science made commercially sourced mobile-location records searchable for law enforcement and how investigators used the system in homicide and missing-person investigations. The Associated Press reported in 2022 that Fog had roughly 40 contracts with nearly two dozen agencies and could search hundreds of billions of records associated with about 250 million devices.² The Electronic Frontier Foundation, reviewing public-records material, documented both “area” searches—find the devices inside a place and time—and “device” searches that reconstruct a particular advertising identifier’s movement history.³

That core account remains important, but the legal landscape has changed substantially. In June 2026, the U.S. Supreme Court decided Chatrie v. United States and held that police conducted a Fourth Amendment search when they acquired a user’s Google Location History through a geofence process. The Court did not decide whether the particular warrant was sufficiently supported by probable cause and particularity; it sent those questions back to the Fourth Circuit.⁵ The decision is highly relevant to location surveillance, but it is not a direct ruling on Fog Reveal or on every purchase of commercially available broker data.

How a location signal becomes a searchable history

The cleanest way to understand the system is to separate the layers. Your phone can determine location using GPS, Wi‑Fi networks, Bluetooth beacons and cellular infrastructure. Apps may request access to that location for a legitimate feature—navigation, weather, nearby restaurants, delivery, exercise tracking—or for advertising and analytics. A mobile advertising identifier (often called a MAID) or other pseudonymous identifier can then connect repeated observations to the same device.

From there, data can enter a complicated commercial chain. Some collection occurs through software development kits embedded in apps. Some data can be transmitted through advertising systems and real-time bidding. Brokers buy, aggregate, enrich and resell data to other companies. The exact lineage can be opaque even to app developers, and investigations have shown that a single device’s data may pass through several intermediaries before reaching a downstream customer.

“Anonymous” does not mean hard to identify

Location data is unusually revealing because routine behavior is repetitive. A device that repeatedly spends nights at the same address and weekdays at another can often be associated with a home and workplace. A landmark 2013 study in Scientific Reports analyzed 15 months of mobility data from 1.5 million people and found that four randomly chosen spatiotemporal points were enough to uniquely characterize 95% of the traces.⁹ The study did not examine Fog specifically; its importance is the general mathematical insight that movement patterns are inherently distinctive.

Real-time bidding multiplied the number of recipients

Real-time bidding (RTB) is the automated auction used to decide which ad is shown when a page or app loads. In a 2022 analysis, the Irish Council for Civil Liberties estimated that the average U.S. internet user’s online activity and location were exposed 747 times per day through RTB, with U.S. data broadcast about 107 trillion times per year. ICCL explicitly described those figures as a low estimate because its underlying industry data did not include all major platforms.¹⁰ These numbers should therefore be read as a dated 2022 estimate of adtech exposure, not as a precise measurement of every person’s phone in 2026.

Why Fog Reveal became attractive to police

Traditional location surveillance often starts with a known person or device: investigators identify a suspect and then seek records associated with that target. Fog-style area searching reverses the sequence. The place and time are known first; the system then surfaces devices that were nearby. This can be useful when investigators have a crime scene but no suspect, especially in violent-crime, missing-person or child-exploitation investigations.

Fog has argued that it supplies commercially available “lead data” rather than evidence that should itself establish probable cause. Public statements reported by AP also emphasized the needs of underfunded departments and the use of the tool in time-sensitive cases.² Privacy advocates answer that the same capability can sweep in people who were merely nearby and can reveal visits to homes, clinics, places of worship, protests, lawyers and workplaces. Those two realities—investigative utility and unusually broad collateral exposure—are the central policy tension.

The Fourth Amendment problem: old doctrine, new data

The Fourth Amendment protects against unreasonable searches and seizures and requires warrants to be supported by probable cause and to particularly describe the place to be searched and the things to be seized. The hard question is not whether location information can be sensitive; modern Supreme Court cases repeatedly recognize that it can. The hard question is when government acquisition of data held by somebody else counts as a “search” in the first place—and what legal process is then required.

The third-party doctrine

Two twentieth-century cases supplied the traditional rule. In United States v. Miller (1976), the Court held that bank records held by banks were not protected by the customer’s Fourth Amendment expectation of privacy. In Smith v. Maryland (1979), the Court held that installing a pen register at a telephone company to record dialed numbers was not a search because the caller had conveyed those numbers to the phone company. Together those cases became known as the third-party doctrine.

The doctrine was developed in a world of limited, purpose-specific business records. Smartphones changed the scale. A modern device can create a persistent, automated chronicle of movement without a user taking any conscious step beyond carrying it.

Carpenter: the first major break

In Carpenter v. United States (2018), the Supreme Court declined to extend the traditional third-party doctrine to the historical cell-site location information at issue. The government had obtained 12,898 location points covering 127 days. The Court held that acquiring the records was a search and that the government generally needed a warrant supported by probable cause, while stressing that the decision was narrow and did not decide every other kind of business record or collection technique.⁴

That last limitation is exactly where commercial location-broker products became legally significant. Agencies and vendors could argue that the government was not compelling a carrier to turn over protected records; it was buying a product already offered in a commercial market. Privacy advocates responded that constitutional protection should not turn on whether the government pays a broker instead of serving legal process on the original collector.

Chatrie in 2026: a geofence acquisition is a search

The Supreme Court revisited digital location privacy in Chatrie v. United States. Police investigating a 2019 Virginia credit-union robbery obtained a geofence warrant directing Google to provide Location History records for phones near the scene. The process began with anonymized device data and narrowed toward identifying information.

On June 29, 2026, the Court held that police conducted a Fourth Amendment search when they acquired Chatrie’s location data because an individual has a legitimate expectation of privacy in cell-phone location information. The Court deliberately stopped there. It remanded for the Fourth Circuit to decide whether the multi-step warrant satisfied probable-cause and particularity requirements.⁵

But Chatrie does not say that every location query is unconstitutional, and it does not directly adjudicate Fog Reveal. The Court itself noted that Google changed Location History in a way that largely foreclosed the geofence procedure at issue. The decision is therefore a powerful doctrinal signal rather than a universal rule for every broker database.⁵

The court cases shaping location privacy

What Chatrie may mean for Fog Reveal—and what it does not

The argument for extending Chatrie to broker data is straightforward: location history can reveal the same intimate facts regardless of whether the government gets it from Google, a wireless carrier, or a broker. If constitutional protection follows the sensitivity and comprehensiveness of the location record, then changing the vendor should not erase the privacy interest.

The counterargument focuses on acquisition method and the Court’s caution in Carpenter. Commercially available datasets may differ in precision, duration, completeness, user consent, source and accessibility. A government agency buying a product on terms also available to private customers is not identical to compelling a service provider to search its records. Courts must still decide how much those distinctions matter.

The regulatory shift: the FTC targets sensitive location-data sales

While constitutional litigation asks what the government may acquire, consumer-protection enforcement asks whether brokers lawfully obtained and sold the data in the first place. That second front became much more significant in 2024 and 2025.

In December 2024, the Federal Trade Commission announced an action against Gravy Analytics and its subsidiary Venntel. The FTC alleged that the companies unfairly sold sensitive location data, including information capable of revealing visits to health-related locations and places of worship, and that they collected or used location data without verifiable consent for commercial and government uses. The complaint also alleged that the companies claimed to process more than 17 billion signals from around one billion mobile devices daily.⁸ In January 2025, the FTC finalized its order, restricting the sale, disclosure or use of sensitive location data subject to specified exceptions and requiring a sensitive-location program.⁷

The 2025 Gravy breach showed the security risk of concentration

On January 4, 2025, Gravy Analytics discovered unauthorized access to its AWS cloud storage environment. The company said an intruder copied data and briefly posted a subset on a dark-web forum; it described the licensed data as consisting mainly of mobile advertising IDs, latitude/longitude and timestamps.¹¹ A threat actor separately claimed to have exfiltrated 17 terabytes. Kroll later documented that claim while analyzing the physical-security implications of geolocation-data breaches.¹² The 17-terabyte figure should therefore be treated as the attacker's claim, not as a company-confirmed total.

The lesson is broader than one incident. A database valuable enough for advertisers, government investigators or intelligence users is also valuable to criminals and hostile intelligence services. Centralized stores of precise movement history can create security risk even if direct names are absent.

Federal policy has moved, but no single comprehensive rule closes every gap

The Office of the Director of National Intelligence issued a framework in 2024 for intelligence-community handling of commercially available information and later updated it as an Intelligence Community Policy Memorandum. The framework recognizes that commercially available datasets can reveal sensitive personal details while still allowing intelligence agencies to access such information under governing law and policy.¹³

The Justice Department separately implemented a national-security Data Security Program in 2025 that restricts certain transactions giving countries of concern access to bulk U.S. sensitive personal data, including geolocation data. That program is aimed principally at foreign-adversary access; it is not a general domestic warrant rule for U.S. law enforcement.¹⁴

Congress has repeatedly considered legislation aimed at the data-broker loophole. In 2024 the House passed the Fourth Amendment Is Not For Sale Act, but it did not become law in that Congress. In the 119th Congress, H.R. 7816—the Protect Liberty and End Warrantless Surveillance Act of 2026—was introduced on March 5, 2026. Title II is labeled the “Fourth Amendment Is Not For Sale Act” and would restrict law-enforcement and intelligence acquisition of certain data-broker records. As of the official GovInfo record available for this article, the bill remained an introduced House measure referred to the Judiciary and Intelligence committees; it should not be described as enacted law.¹⁵

Practical privacy: what users can and cannot control

Individual settings matter, but they do not solve the structural problem. Users can deny unnecessary precise-location permissions, choose “while using the app” instead of “always,” periodically reset or restrict advertising identifiers where supported, remove apps they do not trust, disable background location when it is not needed, and review platform privacy dashboards. Those steps reduce the amount of location data exposed through some channels.

They do not guarantee invisibility. Phones still interact with cellular and Wi‑Fi infrastructure; apps can infer location in multiple ways; advertising and analytics systems may use identifiers other than a single advertising ID; and data already transmitted can persist in downstream systems. Apple’s App Tracking Transparency and similar controls constrain particular forms of cross-app tracking, but they are not a universal deletion switch for every data source, broker or prior record.

The unresolved question

The most important question is no longer whether a smartphone can reveal a person’s life. That is established. The question is how constitutional and statutory safeguards should attach when the same sensitive record moves through private commercial hands before the government sees it.

Carpenter said the third-party doctrine has limits when digital records become a comprehensive chronicle of physical movement. Chatrie said government acquisition of the Google location-history data at issue is a Fourth Amendment search. The Fifth Circuit has treated a Google geofence warrant as an unconstitutional general warrant in the case before it, while recognizing the good-faith exception for the officers who relied on it. The FTC has attacked location-broker practices from the consumer-protection side. Congress continues to debate whether agencies should be able to buy data they would otherwise need legal process to compel.

Fog Reveal sits at the intersection of all four developments. The technology shows what becomes possible when advertising infrastructure is repurposed for investigation. The courts are now deciding which constitutional principles follow the data as it changes hands.