Key related concepts
CO-TRAVELER and Mobile Location Tracking
CO-TRAVELER and Mobile Location Tracking is one of the clearest windows into how mobile metadata became an intelligence instrument.
It matters because it sits at the intersection of four worlds:
- cellphone infrastructure,
- location metadata,
- target development,
- and the global reach of foreign signals intelligence.
This is a crucial point.
CO-TRAVELER was not primarily about listening to conversations. It was about turning the ordinary location traces created by mobile phones into a system for discovering who moved with whom.
That is why this entry matters so much. It preserves the story of how NSA used mobile-network metadata to map movement, infer relationships, and treat intersecting travel patterns as intelligence leads.
Quick profile
- Topic type: declassified location-analytics program
- Core subject: an NSA system for analyzing mobile location metadata to identify intersecting movements and unknown associates
- Main historical setting: documented capability in 2006, analytic formalization by 2012, and public exposure in late 2013
- Best interpretive lens: not “a phone call wiretap program,” but evidence for how location metadata alone can become a powerful relationship-mapping system
- Main warning: the broad architecture is well supported, but many present-day operational details remain classified and the public record is strongest on the 2006–2013 disclosure era
What this entry covers
This entry is not only about one codename.
It covers a mobile-analytics architecture:
- what CO-TRAVELER was,
- what kind of data it used,
- how GSM geolocation worked in practice,
- how co-travel analytics identified unknown associates,
- why the program was treated as foreign-intelligence collection,
- what the 2010-2011 domestic pilot revealed,
- and why the exposure of the program mattered so much.
That includes:
- the 2006 GSM Classification Guide,
- the 2012 Summary of DNR and DNI Co-Travel Analytics,
- the December 2013 Washington Post disclosure,
- the role of EO 12333,
- the huge scale of record intake,
- and the separate controversy over U.S. location-data pilot testing.
So the phrase CO-TRAVELER and mobile location tracking should be read broadly. It names both a collection problem and an analytic solution.
What CO-TRAVELER was
CO-TRAVELER was an NSA analytics toolkit built to work with cellphone location metadata.
The public record makes clear that the program’s purpose was not merely to display where one phone had been. It was to help analysts identify unknown associates of known targets by looking for devices whose movements intersected in meaningful ways.
That matters because the analytic logic is different from ordinary surveillance.
A traditional target may already be known. A co-travel system tries to identify the people around that target:
- who repeatedly appears nearby,
- who moves with them,
- who arrives and departs in synchrony,
- and who can therefore be treated as worthy of further attention.
This is one of the most important ideas in the whole story.
The basic intelligence logic
The strongest public description comes from the Washington Post’s reporting in December 2013.
The paper said the NSA was collecting vast numbers of cellphone location records and using tools collectively called CO-TRAVELER to search for unknown associates of intelligence targets by tracking people whose movements intersected. That framing matters because it captures the program’s strategic purpose.
This was not simply a database of phone positions. It was a machine for turning movement into network inference.
That is why CO-TRAVELER belongs near the center of modern SIGINT history.
Why location metadata was so powerful
Location metadata can reveal a great deal even without content.
If a phone appears near a known target once, that may mean little. If it appears near the same target repeatedly:
- at the same offices,
- on the same trips,
- during the same meetings,
- or on the same border crossings, then a relationship can be inferred even without a single recorded conversation.
This matters because location metadata compresses many forms of intimacy and association into analyzable patterns.
That is what makes CO-TRAVELER historically important. It shows how movement itself became intelligence.
GSM location before smartphone GPS
A common misconception is that all location tracking in this era depended on smartphone GPS.
That is too simple.
The 2006 GMS Classification Guide shows that NSA’s GSM geolocation capabilities were built on network-derived location information such as:
- VLR Global Title,
- Location Area Code,
- Cell ID,
- and sometimes latitude/longitude in reporting.
This is a crucial point.
CO-TRAVELER belongs to the era when mobile-network signaling itself—especially GSM tower relationships, roaming updates, and associated metadata—could be turned into a practical geolocation resource.
The 2006 capability baseline
The GMS Classification Guide is especially important because it provides a baseline for what NSA said it could do before CO-TRAVELER became public.
The guide states that NSA could:
- target, collect, and process GSM details,
- collect GSM calls through the global telecommunications infrastructure,
- process unencrypted GSM,
- process some encrypted GSM depending on the cryptovariable,
- and report the locations of GSM users at several levels of specificity.
This matters because it shows that CO-TRAVELER did not appear from nowhere in 2013. The underlying location and GSM exploitation capability was already documented internally years earlier.
Cell towers as geolocation reference points
The same 2006 GMS guide contains one of the clearest statements in the whole archive.
It says GSM cell towers can be used as a physical-geolocational reference point in relation to a GSM handset of interest and suggests portraying that relationship in a 2–3 kilometer ellipse estimate, depending on environmental conditions.
This is historically important.
It means the system did not need perfect pinpoint GPS to be useful. Tower-based reference data was enough to support mobility and proximity analysis at scale.
That is exactly the kind of input a co-travel system can exploit.
The 2012 analytic overview
By 1 October 2012, the program had advanced enough that NSA produced an internal study titled Summary of DNR and DNI Co-Travel Analytics.
The public snippet from that document says it overviews key elements of co-traveler analytics both under development and operational at NSA. That phrase matters.
It shows that CO-TRAVELER was not a distant research concept. It already existed partly in working form.
This is one of the strongest pieces of evidence that the system had matured into an operational analytic toolkit before the public learned its name.
How the co-travel analytic worked
Public snippets from the 2012 overview are revealing.
They describe GSM analytics using GCID information to identify co-travelers. If two selectors are seen at the same GCID around the same time, they are treated as co-travel candidates.
This matters because it shows the logic in its simplest form:
- shared space,
- shared time,
- repeated enough to matter,
- then elevated into association.
That is the analytic heart of the program.
Why “candidate” matters
The word candidate is important.
CO-TRAVELER did not prove a relationship by itself. It generated leads.
This is historically significant because it places the program inside the larger world of intelligence development rather than courtroom-quality certainty. The system helps analysts decide where to look next.
That distinction matters. It explains why the program could be attractive even if location estimates were imperfect. An imperfect lead can still be operationally valuable.
DNR and DNI metadata
The 2012 overview also matters because it connects CO-TRAVELER to both DNR and DNI metadata.
That indicates the program was not built around only one kind of mobile or internet-derived record stream. It was part of a broader metadata-processing environment.
This is important because it helps explain the scale later reported in the press. CO-TRAVELER was not working from a tiny case file. It was drawing from very large metadata flows and then applying relationship logic to them.
The scale of intake
The most famous number in the public story came from the Washington Post: the NSA was reported to be collecting nearly 5 billion cellphone location records a day.
The same report said the records fed a vast database storing information about the locations of at least hundreds of millions of devices. That made the implications much easier for the public to grasp.
This matters because scale is central to the story.
A handful of location records is tracking. Billions per day becomes infrastructure.
Why the program gathered so much data
The Post report also explained the strategic reason for bulk intake.
Officials said the agency collected location data at scale because its most powerful analytic tools needed large volumes in order to discover unknown associates of known targets. That is one of the clearest explanations of why intelligence agencies collect more metadata than they can immediately exploit.
The answer is simple: pattern discovery requires population-level context.
This is a crucial point.
CO-TRAVELER works best when it can compare many devices, many places, and many overlaps.
Global cables and mobile networks
The public record says much of the location data came from tapping the cables that connect mobile networks globally.
That detail matters because it shows how location surveillance and telecom interception overlapped. NSA was not only watching phones from the air or through domestic carrier requests. It was also harvesting mobile-network data from international communications infrastructure.
This is historically important.
The program fits the larger pattern of post-9/11 SIGINT: find global chokepoints, collect metadata at those chokepoints, then build analytics on top of the result.
Foreign focus and EO 12333
The public justification for the program emphasized foreign intelligence.
The Washington Post said officials described the collection and analysis as lawful and intended strictly to develop intelligence about foreign targets. EFF and other critics argued that the likely governing framework was Executive Order 12333. NSA’s own public EO 12333 page says the order is the foundational authority for NSA’s foreign signals intelligence activities, especially collection on foreign persons wholly outside the United States.
This matters because CO-TRAVELER’s public record fits much more naturally into the EO 12333 world than into the domestic Section 215 phone-records program.
That distinction is essential.
Why foreigners and Americans were not fully separable
Even though the foreign-intelligence framing was central, the program still raised concerns about Americans.
The Washington Post said the NSA did not target Americans’ location data “by design,” but acquired substantial domestic cellphone information incidentally, especially because U.S. phones also travel internationally and use the same global networks.
This is one of the most important complications in the story.
A program aimed outward can still gather inward when the infrastructure and the people using it cross borders. That is one reason location metadata became such a privacy flashpoint.
The separate U.S. pilot
Another major part of the history is the separate domestic pilot that came to light before the main CO-TRAVELER story fully broke.
The Los Angeles Times reported that in 2010 and 2011 the NSA collected sample records showing where Americans were when they made mobile phone calls in order to test how such data could be obtained and processed in bulk. Keith Alexander said the test-run data were never available for intelligence analysis purposes.
This matters because it shows that institutional interest in location bulk collection did not stop at purely foreign use cases.
Even if the pilot was not operationalized for intelligence analysis, it revealed policy curiosity and technical ambition.
Why the U.S. pilot changed the debate
The domestic pilot mattered because it undercut simpler public defenses.
If the agency had experimented with bringing U.S. cellphone location information into its bulk systems, then the line between foreign program and domestic possibility looked less stable. That is why the location controversy intensified so quickly in late 2013.
This is historically significant.
CO-TRAVELER became not only a foreign-surveillance story, but also a story about how close the U.S. government had come to broader domestic location ingestion.
Metadata, not content
A core reading rule for CO-TRAVELER is that it was primarily about metadata.
That means:
- location events,
- cell associations,
- registration details,
- tower relationships,
- and timing patterns, rather than the audio content of calls.
This matters because people often hear “surveillance” and imagine only content interception. CO-TRAVELER shows how much power exists on the metadata side.
Location metadata can reveal:
- meetings,
- routines,
- romantic patterns,
- political gatherings,
- visits to doctors or religious institutions,
- and travel networks.
That is why critics treated the program as deeply intrusive even without content capture.
Co-travel as relationship mapping
In practical terms, the most powerful feature of the program was relationship mapping through mobility.
Who spends time in the same places as a known target? Who departs when they depart? Who repeatedly appears in the same travel corridor? Who is at the same tower cluster just before and after key events?
This is one of the reasons CO-TRAVELER mattered so much to target development.
It let analysts move from a known selector outward into a social and physical network that might otherwise remain invisible.
Why this mattered to intelligence agencies
From an intelligence perspective, the attraction is obvious.
Many targets try to avoid saying incriminating things on the phone. But they still move. They still meet. Their phones still register. Their patterns still leave traces.
That is why mobile location metadata became a prized resource. It offered another way to discover networks even when content was sparse, encrypted, or operationally careful.
Why this mattered to civil-liberties critics
From a civil-liberties perspective, the danger was just as obvious.
Location tracking can reveal not just communication but presence. It can show:
- who was in the same place,
- how often,
- and when.
Critics argued that this threatened freedom of association and privacy in ways ordinary wiretap law had not fully absorbed. That is why organizations like EFF and ACLU reacted so strongly once the program became public.
CO-TRAVELER looked like a machinery of social graphing built from physical movement.
Public exposure in 2013
The decisive public moment came in December 2013 when the Washington Post published the first major account of the program. That disclosure transformed a hidden internal toolkit into one of the most widely discussed symbols of location surveillance in the Snowden archive.
This matters because the public could instantly understand the stakes.
A mobile phone is carried everywhere. A program that maps the movements of hundreds of millions of phones is therefore a program that maps a great deal of human life.
That intuitive clarity is what made CO-TRAVELER so explosive.
Why this belongs in the NSA section
This article belongs in declassified / nsa because CO-TRAVELER is one of the clearest examples of how NSA turned metadata into target-development power.
It helps explain:
- how mobile-network location data could be collected and processed,
- how co-travel analytics worked,
- how foreign-intelligence logic and incidental U.S. exposure could coexist,
- and why movement data became one of the defining surveillance issues of the Snowden era.
That makes CO-TRAVELER more than a headline. It is a structural intelligence-history case.
Why it matters in this encyclopedia
This entry matters because CO-TRAVELER and Mobile Location Tracking preserves one of the strongest examples of movement-based intelligence in the modern archive.
Here CO-TRAVELER is not only:
- a codename,
- a metadata program,
- or a Snowden-era scandal.
It is also:
- a GSM geolocation system,
- a co-travel relationship engine,
- a foreign-intelligence target-development tool,
- a case study in the power of metadata without content,
- and a reminder that the most revealing surveillance sometimes follows where people go rather than what they say.
That makes CO-TRAVELER indispensable to a serious declassified history of NSA programs.
Frequently asked questions
What was CO-TRAVELER?
CO-TRAVELER was an NSA analytics toolkit used to work with cellphone location metadata and identify phones or selectors whose movements intersected with known targets.
Did CO-TRAVELER collect call content?
The public record describes it primarily as a location metadata and relationship-analysis program, not as a content-intercept program.
Was the program based on GPS?
Not mainly in the public record. The strongest leaked documents point to GSM and mobile-network metadata such as cell towers, location areas, VLR information, and related signaling data.
What does “co-travel” mean in this context?
It means that two devices or selectors are seen moving together or appearing in the same places at the same times often enough to make them candidates for a meaningful relationship.
How much data was involved?
The Washington Post reported that the NSA was gathering nearly 5 billion cellphone location records a day and storing location information for at least hundreds of millions of devices.
Was CO-TRAVELER supposed to target Americans?
Public officials said the main collection was intended for foreign intelligence and not designed to target Americans. But reporting also said the NSA incidentally acquired substantial location information involving U.S. devices and travelers.
Did NSA ever test bulk U.S. cellphone location collection?
Yes. Keith Alexander acknowledged that in 2010 and 2011 the NSA tested ingestion of U.S. cellphone location data, but said the test-run records were never made available for intelligence analysis.
Why did the program matter so much?
Because location metadata can reveal meetings, routines, and hidden relationships even without message content. CO-TRAVELER showed how those patterns could be analyzed at global scale.
Related pages
- FASCIA Mobile Location Metadata Database
- BOUNDLESSINFORMANT Data Visualization Program
- EO 12333 and Global Signals Intelligence
- BLARNEY Telecom Intercept Program
- STORMBREW Upstream Collection Program
- XKEYSCORE Search and Query System
- Section 215 Telephony Metadata Program
- SKYNET Predictive Mobile Analytics
- Government Files
- FOIA Releases
- Legal Frameworks
- Congressional Records
Suggested internal linking anchors
- CO-TRAVELER and Mobile Location Tracking
- CO-TRAVELER explained
- NSA cellphone location tracking
- co-travel analytics and mobile metadata
- GSM geolocation at NSA
- CO-TRAVELER and FASCIA
- foreign mobile-location intelligence
- movement-based relationship analysis
References
- https://www.washingtonpost.com/world/national-security/nsa-tracking-cellphone-locations-worldwide-snowden-documents-show/2013/12/04/5492873a-5cf2-11e3-bc56-c6ca94801fac_story.html
- https://www.theguardian.com/world/2013/dec/04/nsa-storing-cell-phone-records-daily-snowden
- https://www.eff.org/deeplinks/2013/12/meet-co-traveler-nsas-cell-phone-location-tracking-program
- https://www.aclu.org/nsa-documents-released-to-the-public-since-june-2013
- https://www.aclu.org/sites/default/files/assets/2013.12.10_cotraveler_overview.pdf
- https://www.aclu.org/files/natsec/nsa/20140130/2013.12.10%20GMS%20Classification%20Guide.pdf
- https://www.aclu.org/files/assets/FASCIA%20Powerpoint%20Slide.pdf
- https://www.latimes.com/nation/la-xpm-2013-oct-02-la-na-nsa-surveillance-20131003-story.html
- https://www.nsa.gov/Signals-Intelligence/EO-12333/
- https://www.eff.org/deeplinks/2013/10/nsa-tracked-americans-cell-locations-two-years-senator-hints-theres-more
- https://www.aclu.org/news/national-security/it-sure-sounds-nsa-tracking-our-locations
- https://www.documentcloud.org/documents/888734-cotraveler-tracking-redacted/
- https://www.dni.gov/files/documents/RG/2013%2010%2004%20BORDC%20memo%20to%20GSI%20review%20group%20--%20FINAL.pdf
- https://www.brennancenter.org/sites/default/files/publications/Overseas_Surveillance_in_an_Interconnected_World.pdf
Editorial note
This entry treats CO-TRAVELER not as a generic cellphone-spying headline, but as a system for turning mobility into analysis. The strongest way to read the program is through proximity. Phones do not merely communicate. They register, roam, attach to towers, move through switching systems, and leave behind traces of place and time. CO-TRAVELER mattered because it treated those traces as social evidence. Once that analytic logic became visible, the privacy stakes were obvious. A government that can reliably see who moves with whom does not need to hear every conversation to learn a great deal about human relationships. That is why CO-TRAVELER matters. It exposed the intelligence value of movement itself.