Start Here
Donate

6 — Legal Basis

This section establishes that using Actual Vote is legal in the United States, describes the specific legal cases under which AV operations are conducted, makes explicit what is not legal under the framework, and — new in v2 — addresses the evidentiary status of AV recordings, the legal pathways through which AV evidence can reach courts and regulators, the subpoena question, and the use of public records requests to obtain poll tapes.

The legal basis comes before the methodology and access methods sections (Section 7 and Section 8) because both of those sections assume legality. A reader who is not yet convinced that AV is legal will not engage with the methodology. A reader who understands that AV is legal but doesn’t know what to do with the evidence it produces needs the broader framework that this section provides.

A note on what this section is and isn’t. This section is not legal advice, and America Counts is not a law firm. The section describes the legal framework that AC relies on for its operations, the evidentiary character of the evidence it produces, and the pathways through which that evidence can matter. Readers who need legal advice for their specific circumstances — for example, before mounting a court challenge based on AV evidence — should consult a licensed attorney in their jurisdiction.

6.1 — General Legal Principle

The legal basis for Actual Vote rests on two long-established principles of US law: that voting results are public information, and that photography in public places is generally legal absent specific prohibition. These principles are stable, well-documented in case law and in popular legal references, and have been the foundation of related civic activities — election observation, public records research, journalism — for decades.

Voting results are public information

In the United States, voting results — the totals reported by election officials at every level, from precinct to state to federal — are public information. They are produced by government officials acting in their official capacity, paid for with public funds, and reported to the public through official channels. There is no legal expectation of secrecy attached to vote totals. Anyone is entitled to see them, write about them, analyze them, and republish them.

This is in contrast to several other categories of election-related information that are protected. Individual ballots are protected to preserve voter secrecy — the right to vote in private depends on the inability of anyone else to see how any specific voter voted. Voter rolls in some states are partially restricted, with portions such as residential addresses treated as confidential. Election workers’ personal information is protected in most jurisdictions, particularly home addresses and contact information. But the aggregated totals — the numbers that result from counting many individual ballots — are public. They have to be public for elections to function: voters need to know the results to know who won, election watchdogs need to verify that reported results match underlying procedures, and the public needs to scrutinize the work of election officials. A democracy whose vote totals were secret would not be a democracy at all.

Poll tapes — the strips of paper that voting machines print at the close of polls, showing the totals tabulated by that machine — are one of the earliest physical artifacts in the chain of vote reporting. They are produced by government equipment in the course of an official function. They contain only aggregated totals, not individual ballots. They are, by both nature and function, public information.

When a poll tape is publicly displayed at a polling place after the polls have closed, the public information on the tape is simply being made visible to the public. There is no legal barrier to anyone — voter, candidate, journalist, or curious citizen — looking at it, reading it, taking notes about it, or making a record of it.

The second principle is broader than elections. In the United States, the general rule is that anyone may take photographs of whatever they want when they are in a public place or in a place where they have permission to take photographs, absent specific legal prohibition. This rule has been articulated many times in case law and is summarized authoritatively in Bert P. Krages II’s reference work The Photographer’s Right (2016), which America Counts has relied on as the canonical statement of the principle since the early days of the project.

Krages’s summary establishes that the general rule permits photography in public places; that examples of public places include streets, sidewalks, and public parks; that property owners may legally prohibit photography on their premises but cannot prohibit others from photographing their property from other locations; and that when a property owner tells you not to take photographs while on their premises, you are legally obligated to honor the request.

These are not novel claims. They are well-established interpretations of US constitutional and statutory law that have been tested repeatedly in court. Photographers have the right to photograph what they can see from public places. Property owners have the right to set rules on their own property. Absent specific legal prohibition — such as restrictions on photography of certain government facilities, courtrooms, or schools — the default is that photography is allowed.

Combining the two principles

Putting the two principles together: a publicly displayed poll tape is public information, and photographing a publicly displayed poll tape is a legal act.

This is the foundation. Most of Actual Vote’s operations consist of users going to polling places after the polls have closed, finding posted poll tapes, and video-recording them. Each of these acts — going to a polling place after polls have closed, looking at a posted tape, video-recording it — falls within the general framework of legal activity. There is no special permission required from any election official, no license needed, no fee to pay, and no restriction on what can be done with the resulting recording.

This is not a controversial legal position. It is the same position that supports election observation by candidates and political parties, journalism about election results, public records requests for election materials, and academic research on elections. Actual Vote uses the same principles to enable a specific activity — independent video documentation of poll tapes by ordinary citizens — but the legal foundation is the same long-established framework.

The role of the Krages reference

America Counts has relied on the Krages quote since the early days of the project as the canonical short statement of the general principle. The Photographer’s Right is a small reference card produced by an attorney specializing in photography law; it has been distributed widely to photographers as a quick-reference guide to their legal rights. The 2016 version is the one America Counts cites.

The reason AC uses Krages rather than more elaborate sources is simplicity and clarity: the Krages summary states the general principle in a few sentences in plain language, and it has the authority of being written by a practicing attorney rather than by AC itself. When a curious or skeptical reader asks “is this really legal?” the Krages reference gives a clean answer. Deeper legal questions about specific situations should be referred to a licensed attorney in the relevant jurisdiction, but the Krages summary suffices for the basic principle.

6.2 — Three Legal Cases for Recording Poll Tapes

The canonical AC document Using Actual Vote is Legal identifies three specific cases under which AV operations are legal. Each case has slightly different requirements and applies to slightly different circumstances. The three cases are exhaustive in the sense that any legal Actual Vote recording fits into one of them. Section 7 of this manual extends the three legal cases into a five-method enumeration that adds two additional access methods (FOIA requests, and a catch-all for special arrangements), but every one of the five methods derives from one of these three legal cases.

Case 1: Tapes publicly displayed at a polling place after polls have closed

This is the simplest case and the one that supports the most common Actual Vote workflow. After the polls close on election night, in jurisdictions that require or permit it, one copy of each polling place’s poll tapes is physically posted in a place visible to the public. An AV user goes to the polling place, finds the posted tape, and video-records it.

The legal basis for Case 1 is direct: voting results are public information, photography in public places is legal, and the act of standing in front of a posted poll tape and recording it combines both principles cleanly. No special permission is needed from any election official. The user is engaging in a legal activity in a place where they have a legal right to be — typically a public sidewalk or a publicly accessible area outside or inside a polling place — using ordinary equipment to capture publicly displayed information.

Several specific considerations apply to Case 1.

Polls must be closed. Recording inside or near a polling place while the polls are still open is restricted in most jurisdictions, often by specific statute. Arizona, for example, has a 75-foot limit around polling places during voting hours, within which most political activity and many forms of recording are prohibited. Other states have similar rules. The Actual Vote workflow specifically begins after polls close, which moves the activity outside these “during voting” restrictions. Users should confirm their local rules, but the general principle is: wait until polls have closed.

Tapes must be publicly displayed. Case 1 specifically applies to tapes that the polling place has actually posted in a public location. If the tapes have not been posted — because the jurisdiction doesn’t require posting, because the poll workers chose not to post them, or because the tapes have been removed — Case 1 doesn’t apply, and the user must rely on a different legal case or no recording is possible at that location.

The user must be in a place they have a legal right to be. Most polling places are public buildings — schools, libraries, fire stations — that are publicly accessible during posting hours. Recording from a public sidewalk, parking lot, or building entrance is within the framework. Recording from inside private property without permission is not.

The user must respect property owner instructions. As the Krages principle establishes, when a property owner tells you not to take photographs while on their premises, you are legally obligated to honor the request. This applies to Actual Vote recording. If a poll worker, school principal, or property owner instructs the user to stop recording or to leave the premises, the user must comply. The user can dispute the instruction if they believe it to be incorrect, but the dispute should be conducted through proper channels, not by ignoring the instruction. The recording can typically continue from a different vantage point — a public sidewalk, a parking lot, or another location from which the posted tapes are still visible.

Case 1 supports access method 1 in Section 7: outside polling places on election night.

Case 2: Recording during the tabulation process with explicit permission

Some AV users have been formally appointed as poll observers (also called poll watchers, election observers, or other state-specific names) or as poll workers (also called election inspectors, officers of election, or other state-specific names). These appointments give the user a different kind of access to the polling place: they are inside during the closing procedure, they can observe the tabulation step where poll tapes are printed, and in some jurisdictions they can record what they see during this step.

The legal basis for Case 2 is the appointment itself plus the rules of the appointing jurisdiction. The user is present in the polling place not as a member of the public (which would be restricted during closing procedures) but as an officially-recognized observer or worker with explicit duties and rights. The recording activity is permissible to the extent that the rules of the appointment permit it.

Appointment is required. The user must be formally appointed by the relevant authority — typically a political party (for poll observers) or the local elections office (for poll workers) — and the appointment must be in effect for the specific polling place and the specific election. Appointment requirements vary widely by state. In Texas, poll watchers are appointed by political parties under the Texas Election Code. In Virginia, “Officers of Election” are appointed by Electoral Boards from lists of nominations filed by political parties. In Wisconsin, “Election Inspectors” must be qualified electors of the county and must complete training. The state-specific rules are documented in the state-by-state appendix (B.3).

Local rules govern what the appointee can do. Different states allow different things during the closing procedure. Some states explicitly permit poll watchers to record poll tapes before they are sealed. Others are more restrictive. In Texas, poll watchers can in some circumstances record tapes before sealing; the Texas Election Code is the authoritative source for Texas-specific rules. Equivalent state codes apply in other states. Users planning to use Case 2 should review the relevant sections directly and consult with their appointing party.

The recording must not interfere. Even when permitted, the recording activity must not interfere with the work being done by election officials. The user must respect any instructions about distance, audio, or timing.

Recording outside the scope of the appointment can have consequences. The recording may be considered inadmissible. The user may lose their observer or worker credentials. The user may be asked to leave the polling place. In extreme cases, the user may face legal consequences. When in doubt, the user should not record. The legal risk to the user and the reputational risk to AC are not worth the benefit of one recording.

Case 2 is the earliest chain-of-custody point for poll tape recordings — there is no opportunity for the tape to be altered between the moment it’s printed and the moment it’s recorded. This gives Case 2 recordings special evidentiary weight, though the practical opportunities to use Case 2 are rarer than Case 1.

Case 2 supports access method 2 in Section 7: poll observer or poll worker recording during tabulation.

Case 3: Arrangement with the local elections office to access tapes after election day

The third case covers situations where a user or organized group makes an arrangement with the local elections office to access poll tapes after election day. This is the case used in jurisdictions where tapes are not publicly displayed on election night (so Case 1 doesn’t apply) and where the user has not been appointed as a poll observer (so Case 2 doesn’t apply).

The legal basis for Case 3 is the cooperation of the elections office. The office holds the original poll tapes and the electronic data as records of the election. As public records held by a government office, the tapes are subject to public records laws and can be made available to the public on request. The exact terms of the arrangement vary: some offices allow visitors to come and view the tapes on-site with permission to photograph or video-record them; some offices scan the tapes and send the scans by email or via a file-sharing service; some offices mail physical copies to a requester; and some offices release the tapes only in response to a formal Freedom of Information Act (FOIA) request.

The cooperation level varies enormously by jurisdiction and by the particular officials involved. Some elections offices are eager to cooperate with independent verification efforts; others are reluctant or actively resistant. Some treat poll tape requests as routine; others treat them as unusual and require extensive justification. AC’s experience includes the full range — from the Henderson County Board of Elections, which responded cooperatively and promptly to AC’s inquiries, to jurisdictions that have been unresponsive or obstructive.

Specific considerations for Case 3 include time pressure (if the analysis aims to produce findings before certification, the arrangement must happen quickly), documentation (arrangements should be confirmed in writing), and cost (some jurisdictions charge for copies or staff time).

The Citizens Audit Broward example illustrates Case 3 in practice. In 2020, Citizens Audit Broward (CAB), a local civic group in Broward County, Florida, made arrangements with the Broward County Supervisor of Elections to record poll tapes from two recent elections at the SOE office. The recordings were made, the data was transcribed, and a Comparison Analysis was performed. The CAB experience demonstrated that Case 3 access works in practice and can support a full analysis pipeline.

Case 3 supports access methods 3, 4, and 5 in Section 7: appointment with the elections office, FOIA request for poll tape images, and special-case arrangements.

How the three cases relate to the five access methods

The three legal cases and the five access methods are two views of the same underlying landscape. The three cases describe the legal basis; the five methods describe the operational paths. Each method derives from one of the three cases:

Method 1 (outside polling places on election night) derives from Case 1. Method 2 (poll observer or poll worker recording during tabulation) derives from Case 2. Method 3 (appointment with the local elections office) derives from Case 3. Method 4 (FOIA request for poll tape images) derives from Case 3 as an extension. Method 5 (special cases — mailed tapes, email scans, novel arrangements) also derives from Case 3 as an extension.

Section 7 treats each method in operational detail. The legal basis for each is established here.

6.3 — What’s Not Legal

The legal framework that supports Actual Vote is generous, but it is not unlimited. This subsection makes the limits explicit so that users do not make assumptions that could get them into trouble.

In nearly every US jurisdiction, recording inside or immediately around a polling place while the polls are still open is restricted. The restrictions are intended to protect voter privacy, prevent intimidation, and maintain order during voting. Specific rules vary: many states impose distance limits around polling place entrances (commonly 75 to 100 feet) within which most political activity, including recording, is restricted during voting hours; some states explicitly prohibit cameras inside polling places during voting hours; and even where general recording is not prohibited, recording individual voters in the act of voting is universally prohibited.

Actual Vote operations specifically begin after polls close, which moves them outside these restrictions. Users should never attempt to record poll tapes while voting is still in progress. If users encounter poll tapes during voting hours — for example, partial tapes printed earlier in the day, or zero tapes from the morning — they should not record them.

The Case 2 (poll observer / poll worker) workflow is a partial exception: appointed observers and workers are inside the polling place during closing, which is technically before all activities have ended. But the appointment grants specific access for specific activities, and the activities permitted by the appointment are governed by the local rules of the jurisdiction. Recording outside the scope of those rules is not authorized.

The Krages principle is clear: when a property owner tells you not to take photographs while on their premises, you are legally obligated to honor the request. Many polling places are housed in private buildings — churches, private schools, community centers, businesses. Even when the polling place itself is publicly accessible during voting hours, the building owner retains the right to set rules on their property. If a property owner instructs an AV user to stop recording or to leave the premises, the user must comply.

The recording can typically continue from a different location — a public sidewalk, a parking lot, or another vantage point from which the posted tapes are still visible. Section 7.2 provides specific guidance on handling encounters with property owners and on alternative vantage points.

The closing procedure at a polling place — tabulation, tape printing, packaging, transport to the central office — is a formal process conducted by election officials in their official capacity. AV users are not authorized to interfere with this process. Users may not handle voting equipment, handle ballots, handle poll tapes (except by photographing them after they have been posted), enter restricted areas of the polling place, or engage with election officials in ways that distract or impede them.

The proper role of an AV user is observation and documentation from outside the official process. Case 2 users interact with the closing procedure under the rules of their appointment, not under Actual Vote.

AV users are members of the public engaging in lawful documentation activity. They are not law enforcement, not investigators with subpoena power, and not authorities of any kind. Users may not demand that election officials produce poll tapes that have not been posted, threaten or intimidate election officials, misrepresent themselves as having authority they do not have, enter areas they are not authorized to enter, or refuse lawful instructions from election officials, poll workers, or law enforcement.

The proper response to a situation where an AV user cannot record what they hoped to record is to document the situation as a context submission, leave the premises if instructed, and report the experience to America Counts for follow-up through proper channels. AC has procedures for following up with elections offices about access issues; individual users do not need to and should not try to handle escalations on their own.

An AV recording captures whatever is in front of the camera. Poll tapes are the intended subject, but the recording may also capture poll workers, other observers, voters lingering in the area, or bystanders. Some of these people may appear by name or by face. Some may have personal information visible — clothing, badges, conversation — that identifies them.

AC vets every submission and rejects any that contain personally identifying information of people other than the submitter. Users are also responsible for not deliberately capturing or distributing PII of others. Users should frame recordings to focus on the poll tape rather than on people, avoid recording audio of conversations between others, and not post AV recordings to social media or other public channels until AC has approved them for public display. The full PII guidance is in Section 10.6 (Safety and Privacy).

6.4 — Evidentiary Status of AV Recordings

This subsection — new in v2, drawing from the Legal Pathways for Actual Vote Evidence technical paper — addresses the question that comes up after “is it legal?”: what kind of evidence does an AV recording actually constitute, and what is it good for?

The seven components of an AV evidence package

An Actual Vote evidence package, for any specific finding, consists of seven distinct components, each a different type of evidence from an evidentiary standpoint.

The video recording. A digital video file that captures the visual content of one or more paper poll tapes. The recording is contemporaneous (created at the time the tape was visible to the recorder), GPS-tagged, and timestamped. This is physical/digital evidence in the traditional sense — a thing that can be played back and that depicts what its proponent claims it depicts. Its admissibility analysis tracks the analysis for any video evidence: authentication under Federal Rule of Evidence 901 and treatment as an “original” under Rule 1001(d).

The metadata. Structured data associated with each submission: submission identifier, user account, GPS coordinates, timestamp, geographic auto-population derived from GPS, user-entered precinct identifier, and any classification or tagging from AC’s vetting process. This is structured data evidence, independently useful for demonstrating chain of custody and for cross-validation against external information.

The vetting record. AC’s documented decision that this submission passed vetting — that it was not rejected for PII, unsuitable content, or other reasons. This is business records evidence — records made and maintained in the regular course of AC’s operations — and supports chain of custody.

The transcription. The structured representation of contest, choice, voting-method, and value tuples read from the poll tape, produced by AC’s transcription workflow. This is derived evidence — a structured rendering of the visual content, with the recording itself as the underlying source.

The comparison logic and inputs. The documented comparison logic for the specific jurisdiction and election: mappings, aggregation rules, edge case handling, and the official results dataset in the form it was published. This is methodology documentation — the specification of how the comparison was conducted.

The comparison output. The structured results of running the comparison: per-grain-tuple comparisons with outcome labels and numerical differences. This is derived analytical evidence — the output of the methodology applied to the inputs.

The investigation history. The documented record of any investigation conducted on apparent discrepancies, following the seven-check process. This is investigatory evidence — the record of AC’s diligence in examining what the comparison output revealed.

Surrounding all seven components is the methodology documentation itself — this manual, the Legal Pathways technical paper, and any additional published materials. The documentation is the basis for any expert testimony about the methodology.

What AV evidence proves and does not prove

When properly produced and presented, AV evidence supports a specific factual claim: that on a given date, at a given polling place, the paper poll tape printed by a given voting machine showed specific totals for the contests audited, and that those totals, when aggregated under the comparison logic appropriate for the jurisdiction, match or do not match the officially reported totals for that combination of polling place, voting method, contest, and choice.

That is the affirmative claim. It is a claim about what was on the locally-printed poll tape and about how the locally-printed values compare to the centrally-reported official values. The claim does not extend beyond those two facts.

To be specific about the claim’s limits: AV evidence does not prove that any individual ballot was counted correctly (the methodology operates at the level of tape-printed totals, not individual ballots). AV evidence does not prove that voters’ intent was correctly captured. AV evidence does not prove fraud — even when an apparent discrepancy survives investigation and is escalated as unexplained, the cause of the discrepancy is not determined by AV’s methodology. A discrepancy can be caused by equipment malfunction, software bug, procedural error, training failure, or deliberate manipulation, and distinguishing among those causes is a different investigation. AV evidence does not prove that any specific person engaged in wrongful conduct. And AV evidence does not establish the legal consequences of any finding.

The narrowness of the factual claim is not a weakness. It is the source of the methodology’s structural credibility. Courts generally credit evidence presented with epistemic humility more than evidence presented with overreach. AV’s narrow factual claim, the documented investigation process that supports the claim, and the structural refusal to attribute causes or intentions to any apparent discrepancy are features that make the evidence credible to a serious court precisely because they do not overreach.

Chain of custody

The submission flow — user device, AC backend, vetting, archive, analysis pipeline — is documented at every step, with persistent identifiers, timestamps, and access logs. The video file’s integrity is preserved through the storage layer. GPS coordinates and timestamps provide independent external anchors that can be verified against the known location and time of the polling place’s posting.

Federal courts have repeatedly held that the proponent of evidence need only show a “reasonable probability” that the evidence is authentic; gaps in the chain of custody go to the weight of the evidence rather than its admissibility. AC’s documented chain of custody clears this standard.

Self-authenticating character

Video recordings are a well-established form of evidence with a well-established admissibility framework. AV’s recordings benefit from several features that strengthen authentication: they are contemporaneous, they carry embedded metadata (GPS, timestamp) that can be independently verified, they are published in a public archive where anyone (including opposing counsel) can examine them, and in some cases multiple recordings of the same tape from different angles provide cross-validation.

The CPA-quality argument

A useful frame for thinking about whether AV evidence should be taken seriously, despite no court having yet ruled on it directly, is the analogy to a Certified Public Accountant conducting an audit. The value of a CPA audit does not depend on the CPA having previously testified in court about a similar audit. The value depends on the CPA following an established professional methodology (Generally Accepted Auditing Standards), documenting the work according to professional standards, maintaining appropriate independence, and being prepared to defend the work if challenged.

AC’s methodology has the same structure. The category to which it belongs — batch-comparison auditing of reported tallies, the reported-tally verification layer of the “evidence-based elections” framework (Stark & Wagner 2012) — has more than a decade of peer-reviewed academic publication, statutory implementation in Colorado’s statewide risk-limiting audits (C.R.S. § 1-7-515, since 2017) and in other states, and acceptance in the relevant scientific community of election forensics specialists. AC’s specific implementation applies the same batch-comparison check that risk-limiting audits perform on sampled batches to a particular type of input (citizen-collected video recordings of paper poll tapes) and a particular workflow (AI-assisted human-verified transcription, with comparison conducted at near-100% coverage of submitted recordings). Unlike a risk-limiting audit, AV does not compute or claim a statistical risk limit; for the audited precincts the comparison is exhaustive rather than sampled, so none is needed, and AV’s claims extend exactly as far as its coverage. (The statistical reasoning behind this posture — including the exact contest-level certifications that need no inference at all — is developed in Appendix D: Statistical Inference from AV Results.) The implementation is documented in this manual; it is reproducible from the underlying recordings; it is conducted by an organization with structural independence from the elections offices being audited; and it is published in a public archive that anyone can verify.

The methodology’s reliability follows from the category’s reliability the same way any specific CPA’s audit follows from the reliability of the underlying GAAS methodology. The implementation’s lack of prior court history is a feature of the fact that no AC analysis has yet produced a finding significant enough to drive a court case, not a feature of the methodology itself.

6.5 — Legal Pathways for AV Evidence

Full treatment: this section summarizes a much longer companion analysis, Legal Pathways for Actual Vote Evidence (v2, June 2026), currently a working draft that isn’t yet ready for publication — we’re happy to discuss the ideas on an individual basis (contact us). The paper adds material not yet reflected below: self-authentication of AV records under FRE 902(13)–(14), the amended Rule 1006 route for transcription and comparison tables, the federal 22-month election-records retention requirement (52 U.S.C. § 20701) as the basis for preservation demands, a one-page executive summary for non-lawyers, and three practitioner appendices (a model expert declaration, an evidence-preservation kit, and a day-by-day checklist for the contest window).

This subsection — new in v2, absorbing v1’s Section 13 and drawing from the Legal Pathways for Actual Vote Evidence technical paper — addresses the question that lawyers, candidates, and civically engaged readers ask most: what happens when AC finds something? How can AV findings actually matter?

The four-level impact model

A useful way to organize the different kinds of value that AV produces is the four-level impact model, from the most pervasive and least dramatic to the rarest and most consequential.

Level 1: Transparency. Every recording that AC publishes to its public archive makes a piece of the local-tabulation step of an election visible that was previously opaque. Before AC, the contents of paper poll tapes posted at polling places on election night were technically public information, but in practice they were almost never recorded, almost never aggregated, almost never compared to anything, and almost never accessible to anyone who was not standing in front of the specific polling place at the specific moment the tape was visible. The information existed but was effectively locked away by the friction of access. AC’s archive removes that friction. A recording in the public archive is searchable, retrievable, and verifiable by anyone with an internet connection. Level 1 value accrues to every recording that AC publishes, regardless of whether any specific recording is ever used in any specific analysis.

Level 2: Quality control. This is what happens when AC compares its transcribed values against the centrally-reported official results and finds that they agree. Agreement is not a non-finding; it is a substantive finding that increases public confidence in the count. Most AC analyses produce mostly agreement. When AC publishes a report confirming that transcribed values matched the official results, the confirmation is the kind of value that a routine post-election audit produces: not dramatic, not newsworthy, but real. Level 2 value is the methodology working as intended in the most common case.

Level 3: Escalation. This is what happens when AC finds an apparent discrepancy that survives the seven-check investigation process. When AC has checked everything internal that it can check, has engaged the elections office in good faith, and has not been able to construct an explanation that fits, the discrepancy is treated as a finding worth escalating. Escalation can take several forms: continued cooperation with the elections office to develop additional information, communication with the relevant state-level oversight body, presentation of the finding to journalists, and outreach to affected candidates or campaigns. AC’s posture at Level 3 is studiedly non-adjudicatory: AC describes what it found, documents the investigation, and ensures the finding reaches people whose role is to answer it. AC does not declare elections wrong, does not allege fraud, and does not speculate about causes or intentions. As of this writing, every apparent discrepancy in every published AC analysis has been resolved during the investigation phase, typically through cooperation with the relevant elections office. The first true Level 3 outcome — an apparent discrepancy that survives investigation and is not resolved through cooperation — will be the most significant single event in AC’s operational history.

Level 4: Legal significance. This is what happens when an AC finding becomes the factual basis for a legal proceeding. Legal significance is the rarest level and the most consequential: rare because most AC findings will never reach a courtroom (Levels 1, 2, and 3 absorb most of the work), and consequential because legal proceedings can produce remedies that no other channel can produce.

Why most AV value lives at Levels 1–3

A useful corrective to the framing that “AV is a tool for legal challenges” is the recognition that legal challenges are the smallest part of what AV does. Most of the work happens at the upstream levels and most of the value accrues there. This is not a weakness. A vote-tabulation auditing methodology whose primary value was producing legal challenges would be a methodology that only mattered when something went badly wrong. AC’s methodology matters whether or not anything goes wrong. When the count is correct, the methodology confirms it, and the confirmation is valuable. When the count is wrong, the methodology surfaces the problem and routes it to the people who can fix it. The legal pathway is the backstop for the rare case where the upstream channels do not produce a satisfactory resolution.

The Legal Pathways technical paper walks through six legal pathways in doctrinal detail. In approximate order of practical accessibility:

Regulatory complaints. A regulatory complaint is a formal request that an administrative agency — the state board of elections, the secretary of state’s office, the state attorney general, the federal Election Assistance Commission, or the US Department of Justice — look into something. Regulatory complaints do not require a lawyer, do not require legal standing in any technical sense, do not have a tight deadline, and do not require any specific kind of evidence. They require a credible factual basis and a clear ask. AC, a partner organization, or an individual citizen can file a regulatory complaint based on an AC analysis. Regulatory complaints are the procedurally simplest pathway and the one most likely to produce AV’s first significant Level 4 outcome. The Monmouth County, New Jersey investigation of 2022–2023 — where a vote-counting irregularity led to an attorney general investigation, independent counsel, a detailed investigative report, and ultimately a court order revoking the original certification — illustrates what effective regulatory action looks like.

State election contests. A state election contest is a lawsuit filed in state court by an unsuccessful candidate (or, in some states, by a qualified elector or taxpayer) asking the court to review and possibly set aside the certified results. State contest procedures vary by state but share a common framework: short deadline (typically days to weeks after certification), broad standing granted by statute, expedited proceedings, specific enumerated grounds, and broad remedial authority for the court if the contest succeeds. State contests are the most direct pathway for AV evidence to overturn or modify a certified result, and the most demanding in terms of preparation: the analysis must be substantially complete before the filing deadline, the evidence must clear formal admissibility hurdles, and the materiality math must work (discrepancies must be large enough to affect the outcome).

Recounts. A recount is an administrative procedure conducted by elections officials to re-tabulate ballots. Recounts happen automatically when the margin is below a statutory threshold and can be requested by a candidate in some jurisdictions. AV evidence does not enter the recount itself directly — the recount works on the ballots — but an AC analysis can inform a candidate’s decision to request a recount, can guide which precincts the recount focuses on, and can be used afterward to verify whether the recount resolved the underlying issue.

Public records litigation. Public records litigation compels an elections office to provide records it has refused to provide voluntarily. Every state has a public records law, and paper poll tapes qualify as public records in essentially every state. Elections offices that refuse to provide them can be sued to compel disclosure. Public records litigation is procedurally clean and generally winnable when the records sought clearly qualify as public records and no exemption applies. The strategic value goes beyond the cases that go to court: the credible threat of such litigation often improves voluntary cooperation with public records requests.

Federal civil rights litigation. The most legally ambitious pathway and also the most uncertain. The federal causes of action that could support a vote-counting challenge — Section 2 of the Voting Rights Act, the Equal Protection Clause of the Fourteenth Amendment, the Help America Vote Act, and the Bush v. Gore uniformity principle — are all real but all difficult. The Supreme Court’s January 2026 decision in Bost v. Illinois State Board of Elections substantially improved the standing posture for candidate-led federal litigation, but federal civil rights claims still require showing that what went wrong violated a specific federal right, and AV evidence does not directly establish the kinds of violations that typical federal civil rights claims require. Federal civil rights litigation is a real but distant pathway for AV evidence.

Criminal investigation. Criminal prosecution of election misconduct is governed by federal statutes (most importantly 18 U.S.C. §§ 241 and 242) and by state criminal laws. AC’s evidence is unlikely to be sufficient on its own for a criminal prosecution — criminal cases require proof beyond a reasonable doubt of every element, including specific intent, and AV evidence is about discrepancies in numbers rather than about the persons or processes responsible. But AV evidence can play a meaningful role at the investigation stage, providing the credible factual basis that justifies opening an investigation, issuing subpoenas, and developing additional evidence. Criminal prosecution is the rarest and most demanding Level 4 pathway; the role for AV evidence in it is supporting rather than primary.

Standing and procedural considerations

A practical consideration that cuts across all six pathways is standing: who has the legal right to bring a case? The answer varies by pathway. Regulatory complaints have the broadest standing (anyone can file). State election contests typically limit standing to candidates, political parties, and sometimes qualified electors. Federal civil rights claims are limited to plaintiffs with a concrete and particularized injury (though Bost has broadened this for candidates). Criminal investigations are initiated by prosecutors, not private citizens.

AC’s institutional posture is to produce evidence that others can use rather than to bring legal proceedings itself. AC is a 501(c)(3) nonprofit whose mission is to produce independent evidence of vote reporting accuracy. The decision about whether and how to use that evidence in legal proceedings belongs to the candidates, attorneys, and organizations who are in the best position to evaluate the specific circumstances.

The Legal Pathways for Actual Vote Evidence: An Analytical Framework technical paper is the full doctrinal treatment underlying this summary. It is approximately 33,000 words, written for lawyers, and covers all six pathways in detail with Florida as a worked example. It analyzes the four main evidentiary objections opposing counsel would raise (authentication, best evidence, hearsay, and expert testimony admissibility under Daubert), develops the methodology-category framing that is the paper’s central analytical contribution, and offers eight prioritized recommendations for strengthening AV evidence’s legal posture over time. The paper is a working draft that isn’t yet ready for publication; we’re happy to discuss the ideas on an individual basis — contact us.

6.6 — The Subpoena Question

This subsection — new in v2, drawing from internal AC correspondence about the subpoena issue — addresses a question that comes up among prospective users: can my phone be subpoenaed?

The question

Some prospective AV users have expressed concern that by recording poll tapes and submitting them to AC, they could expose themselves to having their phones subpoenaed if the recordings become part of a legal proceeding. The concern is understandable — people are naturally cautious about anything that could put them in the middle of a legal process they didn’t initiate.

The answer

The legal framework is straightforward. Subpoenas are instruments of existing legal proceedings — they can only be issued in the context of an active investigation, lawsuit, or prosecution. There is no mechanism for a subpoena to be issued in the absence of a controversy. If there is no case pending, there can be no subpoena.

For an AV user’s phone to be subpoenaed, the following chain of events would need to occur: the user would have to record a poll tape that captured evidence relevant to a specific legal proceeding; a case would need to be filed; and the party seeking the evidence would need to convince a court that the user’s recording was material to the case and could not be obtained by other means.

In practice, this chain is extremely unlikely for most users. If a recording reveals a discrepancy that leads to a case, the video recording itself is already in AC’s public archive — the party seeking the evidence can simply download it. The user’s phone would only be relevant if there was a dispute about the authenticity of the specific recording and the phone contained evidence (such as the original file metadata) that the archived version did not. Even in that scenario, the subpoena would be for the recording and its metadata, not for the phone itself.

Realistic risk assessment

No AV user has ever been subpoenaed in connection with an Actual Vote analysis. This is partly because no analysis has produced a finding significant enough to drive a court case, and partly because even if one did, the public archive would provide the evidence without the need to subpoena individual users.

The realistic risk is extremely low. But it is not zero, and users should be aware of it. The User Agreement and the in-app consent flow both disclose this possibility so that users can make an informed decision about participation.

What users should know about their rights

If a user is subpoenaed: they have the right to legal representation; the subpoena must specify what is being requested; they are not required to produce anything beyond what the subpoena specifies; and they can challenge a subpoena they believe to be overbroad or improper.

In the most realistic scenario — a user is asked to confirm that they shot a specific recording — the testimony is minimal and the legal risk is nil. The user is being asked to verify a fact about which there is no real dispute, which establishes a link in the chain of custody of the evidence.

6.7 — FOIA and Public Records Requests

This subsection provides practical guidance for using public records laws to obtain poll tapes. Public records requests are one of the five access methods (Section 7.5) and derive from the Case 3 legal basis established in Section 6.2.

Every US state has a public records law — Florida’s Chapter 119, North Carolina’s Chapter 132, and equivalent statutes in other states. The federal Freedom of Information Act applies to federal records. Paper poll tapes are public records in essentially every state: they are produced by government officials in the course of an official function, maintained by a government office, and not subject to any exemption that would prevent their disclosure.

Template for requesting poll tapes

A straightforward public records request is often the simplest way to obtain poll tapes in jurisdictions where they are not publicly posted on election night. AC recommends a request along the following lines:

The request should specify the jurisdiction and election (e.g., “Catawba County, November 8, 2022 General Election”), the specific records sought (“all voting results poll tapes generated by tabulation machines operated by the [County] Board of Elections upon closure of the election”), and what is not being sought (“we are not interested in zero tapes for this request”). The request should specify the preferred format (“electronic format — PDF is fine — at no charge”) and indicate that time is a factor if the analysis aims to complete before certification.

Jurisdiction-specific variations

The mechanics of public records requests vary by state. Some states have specific timelines within which the office must respond. Some states allow fees for copies or for staff time spent retrieving records. Some states have exemptions that could be invoked (though poll tapes are unlikely to qualify for any standard exemption). Some states have administrative appeal processes for denied requests, and some require judicial action to compel disclosure.

AC’s experience is that most elections offices respond to well-framed public records requests without difficulty. The framing matters: a request that is specific, professional, and reasonable in scope is more likely to get a prompt response than a vague or confrontational one.

Best practices for framing requests

Be specific. Name the election, the jurisdiction, the type of records, and the format. Vague requests (“all election records”) will be delayed or denied.

Be professional. A polite, factual tone produces better results than a hostile one. The elections office staff are public servants doing their job; a request that treats them as partners rather than adversaries is more effective.

Explain the purpose briefly. “Our election integrity team is conducting an independent analysis” is sufficient. A lengthy justification is unnecessary and may raise defensive reactions.

Request electronic format. Electronic copies (scanned PDFs) are faster and cheaper than physical copies. Most offices can provide them.

Follow up if needed. If the initial request receives no response within the state’s legally mandated timeframe, a polite follow-up is appropriate. If the follow-up is also ignored, escalation to a supervisor or to the state oversight body is the next step.

What to do when requests are denied or delayed

Denials of public records requests for poll tapes are rare but do occur. The most common reasons are: the office claims the records are not yet available (because the canvass is still in progress), the office claims an exemption applies (usually incorrectly for poll tape data), or the office simply doesn’t respond.

When a request is denied: ask for the specific legal basis for the denial in writing; consult the state’s public records law to determine whether the denial is valid; consider filing an administrative appeal if the state provides one; and consider consulting with a public records attorney if the denial appears improper.

The strategic context matters. An elections office that refuses to provide poll tapes when asked politely may be more forthcoming when it understands that the requester is prepared to pursue the request through administrative and legal channels. The Legal Pathways technical paper identifies public records litigation as one of the six legal pathways for AV evidence, and notes that the credible threat of such litigation often improves voluntary cooperation.