Fargo, West Fargo & Cass CountyCriminal defense informationCall 701-929-4886
Fargo and North Dakota criminal defense guide

Subscriber Records and Account Attribution

Subscriber records can identify the information supplied when a phone, internet, email, social-media, cloud, or application account was created or…

ElementsPenaltiesEvidenceNext steps

Account names, registration emails, telephone numbers, billing addresses, payment sources, IP sessions, device identifiers, provider certifications, preservation requests, legal process, account control, authorship, and attribution

Subscriber Records and Account Attribution

Subscriber records can identify the information supplied when a phone, internet, email, social-media, cloud, or application account was created or maintained. They may include a name, address, telephone number, email, start date, service type, payment source, session times, and temporarily assigned network addresses. Those fields can connect an investigation to an account. They do not automatically prove who supplied truthful information, who controlled the account at the charged time, who used a particular device, or who authored a disputed message.

Registration is not authorship. Separate the provider’s account record, subscriber or payer, credential holder, logged-in device, network connection, session event, content author, and human actor. The State must support every material link rather than treating an account label as a person.

Subscriber records describe a provider relationship

The account name may be verified, user supplied, inherited, or false

A provider record can accurately show that an account was registered under a particular name without proving that the name is true. Some services verify identity or payment; others permit nicknames, pseudonyms, organizational names, or minimal registration. The defense should learn what the provider required on the creation date and whether later verification occurred.

Contact fields can identify recovery channels rather than the current user

An email address or telephone number may be used for initial registration, multifactor authentication, recovery, alerts, or billing. It may later be replaced, recycled, shared, or lost. Records should include addition, verification, removal, and change dates where available. A current recovery number does not necessarily identify the person who opened or used the account earlier.

A billing address identifies the service or payer, not every user

Residential internet can be billed to one household member while several people use it. A business account can serve many employees. A family wireless plan can include several numbers and devices. An application subscription can be paid through an app store, gift card, prepaid instrument, or another person’s card. Billing evidence is useful circumstantial proof, but its precise proposition should be stated.

Payment records can add a second account layer

A platform may record a card type, last digits, transaction identifier, payer name, billing address, wallet token, app-store receipt, or bank source. The payment provider holds separate records with its own identifiers and authentication. A stolen card, shared family payment method, employer account, or gift purchase can weaken personal attribution. Investigators should not treat a partial card number as a unique person without obtaining the relevant source records.

Start date and service type define only part of the timeline

Section 2703(c)(2) includes length of service and types of service utilized. Account creation, activation, suspension, reactivation, upgrade, cancellation, and deletion may differ. An old creation date may support long-term control, but it does not establish continuous exclusive use. A service-type field may reveal features available without proving that the user employed them.

Provider-assigned identifiers are usually stronger than display names

Numeric user IDs, account numbers, telephone numbers, SIM identifiers, device identifiers, mailbox IDs, and tenant IDs can distinguish accounts whose display names change. The exact identifier used for preservation and production should be retained. A screenshot showing “John Smith” is less precise than the platform’s immutable account ID tied to the requested records.

Subscriber records and content are different statutory categories

Basic account and session records are not the body of an email, message, stored photo, or document. A return may include headers or activity logs without content. Counsel should identify which process produced each field and prevent witnesses from describing metadata as though it revealed the substance of a private communication.

Preservation, subpoena, court order, and warrant do different work

A preservation request freezes existing provider records temporarily

Under 18 U.S.C. § 2703(f), a qualifying provider must preserve records and other evidence in its possession for 90 days, renewable for another 90 days, pending legal process. Preservation does not itself compel disclosure or create records the provider did not possess. The request, acknowledgment, identifier, scope, and renewal dates should be disclosed.

Basic subscriber categories can be obtained through specified process

Section 2703(c)(2) requires disclosure of enumerated subscriber fields when the government uses an authorized administrative, grand-jury, or trial subpoena or another method permitted by § 2703(c)(1). Broader noncontent records may require a § 2703(d) order, warrant, consent, or other qualifying process. State authority and restrictions must also be considered.

A § 2703(d) order uses a relevance-and-materiality standard

Section 2703(d) requires specific and articulable facts showing reasonable grounds to believe the records are relevant and material to an ongoing criminal investigation. That standard is below probable cause. Carpenter v. United States, 585 U.S. 296, 316–22 (2018), held that this order was insufficient for at least seven days of historical CSLI, while expressly preserving conventional applications of the third-party doctrine.

A warrant may combine subscriber, session, and content requests

A provider warrant should identify account, date range, categories, offense nexus, and court authority. The return should permit counsel to determine which fields were supplied as subscriber data, transactional records, stored content, location, or provider-generated logs. A broad production does not eliminate execution-scope or particularity questions.

No-notice and nondisclosure orders are separate from acquisition authority

Section 2703(c)(3) provides that the government need not notify a subscriber when receiving records under that subsection. A nondisclosure order can restrict provider notice under separate authority. Counsel should obtain the order, duration, findings, extensions, and service history rather than infer legality from the provider’s silence.

Provider portals create identifiers and audit evidence

Large platforms receive requests through law-enforcement portals that assign case or request numbers, record submissions, acknowledge preservation, and deliver returns and certifications. Those materials establish chain and scope. Request the portal correspondence, uploads, amendments, rejection notices, download logs, and complete original production rather than only the investigator’s selected pages.

The provider may produce more or less than requested

Retention, account state, encryption, date interpretation, provider policy, or error can affect production. Investigators should segregate material beyond legal scope and seek clarification for missing categories. A cover letter or field guide may contain limitations essential to interpreting the return. Silence in the production does not prove the event never occurred.

Attribution requires proof beyond the account-registration screen

Account creation may predate the alleged user’s control

Accounts are transferred, sold, shared, inherited, compromised, or repurposed. A family member may create an account for another person. A business administrator may create employee access. The relevant history includes creation device, registration address, verification, recovery changes, username changes, linked accounts, and sessions near the charged event.

Credentials identify access capability, not necessarily the actor

A correct password or authentication token permits entry, but passwords are saved, shared, phished, reused, or synchronized. Multifactor approval can strengthen attribution if tied to a controlled device at the relevant time. A long-lived session may continue without a new password entry. The witness should distinguish login, token refresh, and content action.

Shared and managed accounts defeat assumptions of exclusivity

Couples, families, teams, businesses, campaigns, and social-media managers can share credentials or delegate roles. Platforms may record which page administrator acted or may display only the public page identity. Workplace single-sign-on and mobile-device management add organizational control. The account name can remain constant while human users change.

Account compromise is a factual issue, not a magic phrase

Security alerts, unfamiliar sessions, password resets, recovery changes, phishing, malware, new devices, provider reports, and contemporaneous complaints can support compromise. The absence of a report is relevant but not conclusive. A defendant’s bare claim of hacking does not prove it, and a provider’s failure to flag compromise does not prove exclusive control.

Distinctive content can connect an account to a person

Private knowledge, unpublished photographs, recognizable voice, consistent relationships, real-time plans, replies understood by recipients, and later conduct may support attribution under all the circumstances. The supposed distinctive fact should be tested: it may have been public, shared in a group, or available in earlier account content.

Possession of a logged-in device is strong but time specific

A device seized from the accused while logged into the account supports control at seizure. It does not conclusively establish control months earlier. Review session history, device additions, cookies, local databases, notifications, account tokens, and possession evidence at the charged time. Shared or unattended devices may permit another user.

Authorship is narrower than account control

A person can control an account without writing every post or message. Automated scheduling, connected applications, bots, imports, forwarding, dictation, and another authorized user can create content. Trial proof should connect the accused to the specific act, not merely the broader account.

Session and device records can strengthen or weaken attribution

Registration IP and event IP prove different moments

A registration IP may show the connection used when the account was opened. A login or activity IP records a later session or event. Neither necessarily locates the user precisely. Dynamic assignment, mobile gateways, VPNs, proxies, businesses, and shared Wi-Fi require the exact address, time, port where needed, and provider assignment record.

A session time may reflect background activity

Applications refresh tokens, synchronize messages, upload backups, check notifications, and maintain connections without active human use. A provider’s “login,” “session,” or “last active” field must be defined. Device usage, foreground application records, message creation, and content events may be needed to prove contemporaneous action.

Device labels may be generic or user supplied

A security page might display “iPhone,” “Chrome on Windows,” or a custom device name. User-agent strings can be common, altered, or incomplete. More specific identifiers—installation ID, cookie, hardware or advertising identifier, SIM or handset number—have their own reset, sharing, and privacy behavior. A device label should not be called unique without proof.

IP evidence identifies a connection before it identifies a human

A subscriber return can associate an IP session with a household, business, carrier customer, or VPN. Router logs and device forensics may connect that traffic to an internal device. Account tokens and application databases may connect the device to the account. Possession and surrounding evidence then may connect the device to a person. Each inferential step should remain visible.

Location inferences must distinguish service address and network exit

A residential ISP assignment may include a service address. Mobile and VPN addresses may geolocate to a regional gateway. Commercial geolocation is an estimate, not GPS. A map pin should identify the source and uncertainty. Describing a network exit as the defendant’s physical location can mislead the jury.

Change history can reveal takeover or continuity

Username, password, recovery email, telephone, payment, device, and security-setting changes create a control history. A sudden recovery change followed by new sessions can support takeover; consistent long-term sessions and private content can support continuity. The provider may not retain every change, so absence requires a completeness foundation.

Multiple independent sources matter more than repeated derivatives

A platform return, officer spreadsheet, warrant affidavit, and timeline may all reproduce one account record. They are not independent corroboration. Strong attribution combines genuinely separate sources: provider logs, ISP assignments, device artifacts, witness communications, possession, payment, and content. A provenance chart prevents double counting.

Business-record authentication does not establish account authorship

Rule 901 asks whether the exhibit is what its proponent claims

N.D.R.Ev. 901(a) requires sufficient evidence to support a finding of authenticity. A provider custodian, certification, investigator with portal knowledge, unique production identifier, and distinctive record structure can authenticate the production. The evidentiary claim should be precise: provider records for an identified account, not “records proving the defendant sent the message.”

Rule 902 permits self-authentication of qualifying records

N.D.R.Ev. 902(11) addresses certified domestic records of a regularly conducted activity, while Rules 902(13) and (14) address qualifying electronic-process records and copied electronic data. Notice and inspection allow the opponent to challenge certification. Self-authentication removes a foundation witness requirement; it does not resolve hearsay inside the record or identity of a user.

The business-record exception covers provider practice, not every user assertion

N.D.R.Ev. 803(6) can support records made at or near the time by someone with knowledge, kept in regular activity, and made as a regular practice, subject to trustworthiness. Provider-generated timestamps and identifiers differ from a user-entered name, biography, address, or message. Embedded statements need their own analysis.

Azure requires circumstantial evidence linking the account to the defendant

In United States v. Azure, No. 24-2363, slip op. at 9–12 (8th Cir. Jan. 15, 2026), the Eighth Circuit affirmed admission of certified Facebook records from a North Dakota prosecution. The court emphasized that a platform certification alone does not authenticate social-media authorship. Witness testimony, a unique Facebook identifier, recognized friends and communications, legal-process records, and an IP address tied to a relative’s nearby residence collectively supplied a rational basis to send authenticity to the jury.

Henderson warns against foundation by mere receipt

State v. Henderson, 2024 ND 42, ¶¶ 9–27, reversed convictions where central documents, images, receipts, and video lacked adequate authentication or business-record support. An officer’s receipt of an account spreadsheet does not establish how the provider created each field or whether a user-entered assertion is true.

Hearsay purpose and account attribution remain separate

An alleged statement by the accused may qualify under N.D.R.Ev. 801(d)(2) only after sufficient evidence connects it to the party. Incoming messages, profile statements from third parties, and forwarded content can contain additional hearsay. A provider certificate authenticates the recordkeeping process, not the truth of every displayed statement.

Expert testimony must fit the disputed link

Under N.D.R.Ev. 702, specialized testimony must help the factfinder and rest on sufficient facts and reliable methods. A network expert may interpret IP sessions; a forensic examiner may connect device artifacts; a provider witness may explain fields. No witness should leap from a technical record to human authorship without an adequate factual basis.

Common attribution errors collapse several distinct questions

Treating the subscriber as the exclusive user

A parent, employer, landlord, account administrator, or payer may be the named subscriber while others use the service. The report should identify the subscriber relationship and separately prove control at the relevant time.

Treating a registration field as verified identity

A platform can accurately preserve a false or outdated name. Determine whether identity, email, phone, payment, or government document verification occurred and when. A field is not verified merely because it appears in an official return.

Treating account access as message authorship

A background session, shared credential, or synchronized device may appear in logs without a human composing content. Match the content event to session, device, application, and contemporaneous possession. Avoid phrases such as “the defendant logged in and sent” when the record proves only one part.

Treating an IP map as physical-location proof

An IP can associate an account with a service connection or network exit. It generally does not locate a person inside a house or identify a handset. Mobile gateways, VPNs, and geolocation databases require explicit limits. Use subscriber addresses, network estimates, and device coordinates as distinct categories.

Treating the provider production as complete

The request may be date limited; the provider may lack old fields; content may be encrypted; some logs may rotate; linked services may use separate systems. The cover letter, field guide, request language, and record counts define the return. Absence from a production is not absence from reality without a retention foundation.

Treating several copies as independent corroboration

An investigator’s notes, report, warrant paragraph, and demonstrative may all derive from the same provider row. Count original sources, not exhibits. This distinction matters when one error—wrong time zone, mislabeled identifier, or unsupported name—propagates across the case.

Prepaid and minimally verified accounts require different proof

A prepaid telephone, gift-funded application account, or free email address may contain little verified subscriber information. The absence of a conventional billing name does not make attribution impossible; activation location, retailer records, payment instruments, recovery contacts, linked devices, sessions, and content may still supply evidence. The State should not describe an unverified registration field as provider-confirmed identity, and the defense should not imply that a pseudonym alone defeats otherwise strong corroboration.

Recycled numbers and addresses can connect the wrong period to the right person

Telephone numbers, dynamic network addresses, and even email aliases may be reassigned. A current subscriber record does not establish historical control. The return should supply assignment start and end times, and the request should use the exact event time and time zone. Account-recovery records are particularly vulnerable when a number belonged to one person at account creation and another when police later queried it.

Organization-managed accounts separate administrator from end user

Employers, schools, agencies, and families may provision accounts centrally, reset credentials, retain logs, and access stored material. The named administrator may never use the account’s substantive features, while an employee may lack control over billing and recovery fields. Role assignments, directory records, device management, access logs, employment dates, and policy evidence are necessary to identify who could act and who actually did.

Defense counsel should build an account-attribution chain from native records

Obtain every request, response, and identifier

Request preservation letters, renewals, subpoenas, § 2703(d) orders, warrants, affidavits, nondisclosure orders, provider portal receipts, amendments, rejection notices, correspondence, certifications, native returns, cover letters, field guides, download logs, inventories, and derivative warrants. Match each returned field to its legal authority.

Construct an account-control chronology

Record creation, verification, usernames, recovery contacts, telephone numbers, payment, password resets, multifactor changes, device additions, sessions, security alerts, content events, preservation, suspension, deletion, and production. Overlay device possession, residence, travel, witnesses, and alleged offense times. This chronology can show continuity, sharing, or takeover.

Retain native values and provider definitions

Preserve numeric account IDs, raw timestamps, IPs, ports, device IDs, status codes, source tables, and original file formats. Record the provider’s definition for each field and every time conversion. A PDF with friendly labels may hide whether “user,” “subscriber,” “actor,” and “device” mean different things.

Compare the provider return with device and network evidence

Account tokens, application databases, cookies, notifications, browser history, linked emails, saved passwords, device-usage logs, Wi-Fi records, ISP assignments, and router logs can corroborate or contradict provider sessions. A match should be tested at the exact event time. A later logged-in device may not prove earlier use.

Investigate compromise and shared control with evidence

Obtain security emails, reset notices, unfamiliar sessions, phishing messages, malware evidence, remote-access tools, role assignments, household or workplace access, and reports to the provider. Test alternative users’ devices and schedules where lawful. A supported alternative is more persuasive than a generic claim that anyone could have used the account.

File focused motions and make precise objections

Possible issues include overbroad process, missing nexus, acquisition beyond scope, inadequate certification notice, incomplete discovery, unauthenticated user fields, hearsay, misleading location evidence, unsupported expert attribution, Rule 106 completeness, and Rule 403 overstatement. Identify the exact record, missing inferential link, prejudice, and requested relief.

Translate attribution gaps into the elements

Account evidence matters only through the charged offense. If identity is disputed, show why subscriber or payer data does not identify the actor. If knowledge is disputed, show that synchronization or shared storage does not prove awareness. If intent is disputed, examine complete content and conduct. The defense should acknowledge strong corroboration while explaining why the total admissible proof does or does not establish guilt beyond a reasonable doubt.

Frequently asked questions about subscriber records and attribution

What are subscriber records?

They are provider records about an account or service relationship, such as name, address, telephone, email, service dates, payment source, session information, and assigned identifiers.

Do subscriber records prove who used an account?

No. They can identify the registered subscriber or payer. Sessions, devices, credentials, content, possession, and corroborating evidence may be needed to identify the user.

What does § 2703(c)(2) cover?

It lists name, address, connection or session records, length and type of service, telephone or other subscriber identity including temporary network address, and payment source.

Is a preservation request the same as a subpoena?

No. Preservation temporarily requires retention of existing records pending process. It does not by itself compel disclosure.

Can police obtain subscriber information without a warrant?

Federal law permits specified basic subscriber categories through authorized subpoena or other listed process. The exact record, state law, and current constitutional law must be reviewed.

Does Carpenter require a warrant for all subscriber records?

No. Carpenter’s holding concerned at least seven days of historical CSLI and was expressly narrow. It did not categorically invalidate conventional third-party business-record process.

Can a false registration name still be a valid business record?

The provider record may validly show what information was supplied. Business-record status does not establish that the user-supplied name was true.

Does a payment card prove the account owner?

It is circumstantial evidence. Shared, stolen, employer, prepaid, app-store, wallet, or gift payment can weaken attribution and may require separate source records.

Does a login IP locate the user?

It identifies a network connection or intermediary. Provider assignment may identify a subscriber service, but shared Wi-Fi, mobile gateways, VPNs, proxies, and cloud hosts limit physical-location and human inference.

Does a session log prove active use?

Not necessarily. Background refresh, token renewal, synchronization, and automation can create session events. The provider’s event definition and device activity matter.

Can several people use one account?

Yes. Families, businesses, teams, administrators, and shared devices may permit several users. Role and session history should be examined.

How can account compromise be investigated?

Review security alerts, recovery changes, password resets, devices, IP sessions, multifactor records, phishing, malware, provider reports, and conduct around the disputed event.

What does a provider certification prove?

It can authenticate qualifying provider records and establish business-record requirements. It does not prove that the defendant personally used the account or authored content.

What did United States v. Azure decide?

The Eighth Circuit affirmed admission of certified Facebook records where extensive circumstantial evidence linked the account to the defendant. It reiterated that a platform certification alone is insufficient for social-media authorship.

Are user-entered profile fields hearsay?

They may be human statements and require a nonhearsay purpose or exception. A provider business-record certification does not automatically prove their truth.

Can a logged-in phone prove authorship?

It supports account control at a time but does not conclusively prove who wrote every message. Shared devices, sessions, automation, and timing require examination.

When is an expert useful?

An expert may interpret provider fields, IP sessions, device identifiers, forensic artifacts, time conversions, compromise evidence, and the distinction between account and human activity.

What should defense counsel request?

All legal process, portal records, certifications, native returns, field guides, raw timestamps, account and change history, device and session IDs, IP and port data, and derivative exhibits.

Official authorities cited

Legal information, not legal advice. Subscriber-record analysis depends on the provider, account identifier, registration requirements, legal process, preservation, certification, retention, fields, raw times, payment, recovery and change history, sessions, IP and port data, device artifacts, shared access, compromise, authorship, expert foundation, and current law.


Call nowChoose a time