A Statement of Democratic Principles of AUDIT USA

We do not believe democratic societies are defined by perfection, moral purity, or the absence of abuse.

We believe they are defined by their capacity to detect error, correct abuse, and restore trust through evidence.

History shows that when power—whether governmental, institutional, or private—operates without transparency and independent verification, abuse becomes not only possible but predictable. This is not a claim about motives or ideology. It is a structural reality observed across time and systems.

Our work is grounded in three principles:

1. Skepticism is not cynicism.
   Asking for evidence is not an accusation. It is a democratic responsibility.

2. Transparency is not exposure.
   Protecting privacy and civil rights is compatible with public verification of outcomes.

3. Trust must be earned through systems, not personalities.
   Good people operating inside opaque systems are still constrained by what the public can verify.

We do not seek to persuade people to believe us.
We seek to help build systems that allow citizens to verify for themselves.

That is the difference between authority and legitimacy.
And it is the foundation of a durable democratic republic.

Podcast video-Style Report on Cochise County Could Set a New Standard for Election Transparency. (Later I will get the water mark removed)

What Is ABE?

Auditable Ballot Examination (ABE) is a nonpartisan, county-based election transparency method that allows both election officials and the public to independently verify election results—without compromising ballot secrecy or voter privacy.

  

Using records that modern voting systems already generate—ballot images, Cast Vote Records (CVRs), and precinct-level reports—ABE reconstructs election results at the precinct level. This makes it possible for citizens, journalists, candidates, and election officials to confirm that reported outcomes accurately reflect the ballots actually cast.

In essence, ABE brings elections back to a 1912-style precinct-level verification model, updated for the digital age. (See video part of Kent Bennett’s presentation 14 minutes) Instead of relying solely on centralized tabulation and post-election assurances, ABE enables distributed verification using transparent, observable evidence.

ABE does not centralize voter data and does not expose voter identities. Authority remains at the county and state level, while public trust is restored through transparency rather than blind trust in closed systems.

“In short: voting remains secret, counting becomes publicly verifiable, and confidence is earned through observable evidence—not blind trust in a closed process.”

“We anticipate a future in which every precinct committee member and candidate in the county will receive a CVR spreadsheet via email, complete with hyperlinked ballot images accessible for download from the cloud as needed to conduct a risk-limiting audit of their precinct.”
— Ken Bennett

  A Remarkable Meeting

On Thursday, December 4, 2025, at 10:00 a.m., something rare happened in American politics.

In a public hearing room on Melody Lane in Bisbee, Arizona, all key decision makers were present. Members of the Cochise County Board of Supervisors sat alongside the County Manager, the County Recorder, the Elections Director, and senior county staff.

“No yelling. No conspiracy circus. No cable-news theater.”

Instead, leaders of a rural border county—the fifth largest county in Arizona with just over 82,000 registered voters—sat down at a specially called public meeting with two invited guests:

Ken Bennett, former Arizona Secretary of State, and John Brakey, co-founder and Executive Director of AUDIT Elections USA.

They talked calmly, seriously, and publicly about how to let the people verify elections for themselves.

“Somebody’s got to be first. Why not you?”
— Ken Bennett, former Arizona Secretary of State

Video of study session of December 4, 2025:

That challenge framed the entire discussion. Cochise County may now be positioned to become a first mover for a new standard in American elections: radical transparency paired with public verification, using a hybrid system called Auditable Ballot Examination (ABE).

“Apostles for Radical Transparency”

Before coming to Cochise County, Bennett picked up the phone and called Arizona’s Secretary of State to explain what would be presented.

  

There was no attempt to shut the discussion down. The state official didn’t say “back off.” He didn’t say “this will cause trouble.”

Bennett relayed the response he received:

“You can tell them that you and I—Ken and Adrian—are both apostles for radical transparency.”

That line landed in the room like a stake in the ground and was well received with smiles.

This is not a left versus right project. This is not a “who really won 2020?” project.

This is a much more basic—and overdue—question:

Can we verify elections together as adults, using evidence that already exists?

The Man Who Lost Sleep Over 0.001%

Ken Bennett is not just another talking head. He is a native Arizonan, a former city council member, former President of the Arizona State Senate, and he former Arizona Secretary of State from 2009 to 2015. Today, he serves as Chairman of AUDIT Elections USA and remains one of the most credible voices in the country on how to fix the election trust crisis without burning the house down.

He opened the Cochise meeting with a story that still gives him chills.

In 2010, as Secretary of State, Bennett had to oversee the first statewide recount in Arizona history. About two million ballots had been cast. A statewide proposition was losing by just 126 votes.

Under Arizona law at the time, any margin under 200 votes triggered an automatic recount.

Ken Bennett is an accounting major, and as he lay down to sleep that night, he did some quick math in his head.If a recount of two million ballots were:

• 99% accurate, the margin of error could be 20,000 votes

• 99.9% accurate, it could still be 2,000 votes

• 99.99% accurate, it would mean 200 votes — enough to change the outcome

Knowing this, Bennett did not sleep well that night.

After recounting ballots from all 15 Arizona counties, the final result changed by 42 votes. The proposition still lost—now by 168 instead of 126.

More than 1,600 precincts were reviewed statewide. 99.5% of precincts showed no change at all. One precinct in Maricopa County accounted for 30 of the 42 vote shift due to a double counting error when ballots briefly fell behind a tabulator and were run twice.

By accounting for that one mistake, the entire rest of the state only changed by 12 votes out of two million.

The lesson Bennett brought to Cochise was simple and profound:
“Little things matter—especially in very close elections. In small counties, little things are big things.”

And right now, we are living in a world of close elections.The Trust Crisis No One Can Ignore  In recent polling:

• Nearly 4 in 10 Democrats, and
• More than 7 in 10 Republicans say they do not trust the result of at least one of the last two presidential elections.

That is not a fringe problem. That is a regime confidence problem in a nation that calls itself a democratic republic.

We’ve all heard the lines:

• “My vote doesn’t matter.”

• “It’s a rigged game.”

• “If voting really made a difference, they wouldn’t let us do it.”

• “Both sides are corrupt.”

When that attitude settles in, democracy doesn’t just feel shaky—it becomes unstable.

At AUDIT Elections USA, we operate from a simple principle that guided the entire Cochise meeting:
Voting must be secret. Counting must be public.
Trust comes from verification—not faith.

Ken brought up a quote attributed to St. Augustine:

“The truth is like a lion. You don’t have to defend it.
You just let it loose and it will defend itself.”

  How Elections Used to Be Transparent—and How ABE Revives That TraditionOne of the most memorable moments of the Cochise meeting came when Bennett took everyone back to 1912, the year Arizona became a state.In those early elections:

• Ballots were read aloud in each precinct.
• Totals were written on tally sheets in front of observers.
• Each ballot was punched with a hole and slid onto a string.

That string of ballots served as the original public chain of custody. Any citizen—Democrat, Republican, independent, or unaffiliated—could verify that the tally sheet matched the ballots themselves.

Bennett told the room:

“Let me show you how elections were conducted in Arizona in 1912.” It was not nostalgia. It was a lesson.

That same spirit is what animates Auditable Ballot Examination (ABE). ABE is the modern successor to the 1912 system. Instead of paper and string, we now have:
• Digital ballot images
• Cast Vote Records (CVRs)
• Excel based precinct level audit tools

The principle is unchanged: transparency the public can see.

We even named the program ABE as a nod to Abraham Lincoln — a leader who believed deeply in the people’s ability to govern themselves when given the truth.

When one side sees elections as fair and the other sees them as rigged, we’re no longer sharing a democracy—we’re fighting over a broken scoreboard. That’s why tools like ABE exist: to rebuild trust through public verification and transparency.

Former Arizona Secretary of State Ken Bennett: Five Steps to Achieving Verified and Trusted Elections (video)

From 1912 String Ballots to a Digital Hybrid 2026 Canvass System — Simple, Fast, Precise, and provides election directors with Speed and Accuracy.

The core challenge facing election administrators today is that ballots are no longer stored or reported by precinct in a way the public can easily examine.

As Arizona shifted to vote centers and widespread vote by mail, precincts became “virtual.” Ballots now arrive in mixed batches—thousands at a time—from every corner of a county.

This creates a major obstacle for meaningful audits: reassembling a single precinct’s ballots within short legal timelines has become nearly impossible.

In 1912, Arizona candidates had 20 days after certification to gather evidence and challenge an election. Today, the time to challenge an election was reduced from 20 days to just 5.

Yet you cannot challenge what you cannot see.
To inspect original ballots in Arizona, a court order is required, and judges must first see “convincing evidence”—creating a legal Catch 22.

The election document most susceptible to harmful or accidental changes is the Cast Vote Record (CVR)—essentially a large spreadsheet that records how each anonymous ballot was tabulated.

Early ballot processing begins up to 14 days before Election Day. The first one or two days of tabulated ballots—often 20,000 or more—function like a statistically powerful early poll.

(For documented historical context, see Bill Risner’s Statement of Facts in the 2006 Pima County RTA case.)

This potential for errors in the CVR is one of the primary reasons ABE was created.

What ABE Actually Does (In Plain English)

Modern voting systems already create two powerful datasets:
Ballot Images — digital images of every ballot scanned
Cast Vote Records (CVRs) — spreadsheets listing how each anonymous ballot was recorded

ABE simply does three things:

1. Links each ballot image to its corresponding CVR row

2. Organizes the data by precinct using Excel macros

3. Allows humans to verify that the ballot image matches the CVR

At a verification station:

• One person reads the ballot image
• One person reads the CVR row

  

Machines do the heavy lifting. People deliver the truth with facts.
ABE doesn’t replace technology. It proves what the technology did.

Why Counties Like Cochise Can Use ABE Right Now

Cochise County does not need to replace its tabulators or switch vendors.

To use ABE, a county needs:

• A standalone windows computer (offline from the EMS)
• Microsoft Excel
• Ballot images
• The CVR database

That’s it.

Yet despite this readiness, counties face a barrier that has nothing to do with technology—and everything to do with how transparency laws are being interpreted and abused.

The Legal Obstacle to Transparency — and Why It’s Being Misused

One of the central challenges facing election transparency in Arizona is resistance to releasing ballot images—even though they are public records under state law once voter identity has been separated.

That resistance is most often justified by invoking A.R.S. §16-625. To understand why this statute is now being misused, the history matters.

A.R.S. §16-625 did not emerge from a neutral study process, bipartisan reform effort, or a demonstrated privacy failure. It was enacted as a striker bill during active litigation in which citizens were seeking access to ballot images in Pima County. While the case was still before the court, county leadership advanced legislation that changed the rules midstream.

A striker bill is a legislative tactic in Arizona where the original text of a bill is completely removed and replaced with new language—sometimes on an entirely different subject—while keeping the same bill number.

In other words, instead of allowing a judge to rule on whether ballot images were public records—as decades of public-records law would suggest—the Legislature was asked to intervene on behalf of Pima County administrators Chuck Hackelbarry who was actively resisting transparency.

NOTE: Chuck Huckelberry served as the Pima County Administrator from 1993 until his resignation in 2022. He began his career with Pima County in 1974 as a field engineer and advanced through various positions. In 1993, he was appointed as County Administrator, a role he held for nearly thirty years.

Q: Why Transparency?  (For documented historical context, and proof see Bill Risner’s Statement of Facts in the 2006 Pima County RTA case.)That sequence matters.

Striker bills are not inherently illegal, but they are often used when speed and insulation from public scrutiny are the goal. In this case, the effect was immediate: election officials gained a statutory talking point to deny access to records that had previously been treated as public.

The justification offered was ballot anonymity. But anonymity was never shown to be at risk at scale. No evidence was presented that releasing ballot images—without voter-identifying information—had exposed voters or compromised elections anywhere in Arizona.

Instead, the statute created a chilling effect. Counties now routinely cite §16-625 as a blanket prohibition, even though its language does not explicitly bar the controlled release of ballot images for public verification.

This pattern—using privacy rhetoric to block transparency—is not unique to Arizona. But it is particularly consequential here, because Arizona already produces some of the most detailed election data in the country.

Cochise County’s December 4 study session demonstrated that transparency and privacy are not opposing values. When managed correctly, they reinforce one another.

Arizona Case Study: Why the Hand-Count Audit Under A.R.S. §16-602 is now Insufficient

Arizona’s hand-count audit statute, A.R.S. §16-602, no longer functions as a meaningful audit in modern digital elections—not because of misconduct, but because of how ballots are selected, and which contests are excluded.

Under current practice, counties typically begin selecting audit batches 10–14 days before Election Day, while early ballots are still being processed. These selections are described as “random,” but in reality the batches are identified and logged by the tabulation system in advance, even if human observers do not view their contents.

Once those batches are per tagged, meaningful randomness is lost.

From that point forward, the system itself knows which batches are designated for audit and which will never be examined. Any targeted error or manipulation—if it were to occur—would simply avoid those pre-selected batches. The audit then reviews only what the system already protected.

On Election Day, voting-center ballots in Maricopa are tabulated separately. These ballots—approximately 7.5% of total votes—are genuinely unpredictably and therefore the least attractive target for manipulation.

Ironically, this is the only part of the process where randomness protects the audit.

  The greater vulnerability lies elsewhere:

The Cast Vote Record (CVR) database, where results are compiled after ballot images are generated. Without public access to ballot images, there is no independent way to verify that CVRs were not altered after tabulation.

The Cast Vote Record (CVR) where results are compiled after ballot images are generated and interpreted. Without public access to ballot images or the original ballots, there is no independent way to verify that CVRs were not subject to error or altered during or after tabulation.

Why the August 2020 Maricopa Primary Hand Count Wasn’t an Audit  

I witnessed the audit firsthand as a Libertarian Party observer.

The Democratic Party later claimed I “participated in selecting the 25 boxes out of 50.”

Yes, I was there and I was not happy.

There is also another serious, long-standing flaw in Arizona election audits: There are no audits whatsoever for county races, county referenda and county bond issues – they are excluded entirely under Arizona statute §16-602.

For comparison here is what we counted in 2008:

• 2008 audit: 17,075 ballots (with a smaller population)

• 2020 audit: 8,114 ballots (with far more voters)

The required sample size collapsed because the law was written for precinct-based elections, not countywide vote centers. Precinct polling has been replaced by vote centers and vote centers now account for only 7.5% of ballots cast in Maricopa County on election day.

  That exclusion was not accidental.

Former legislator Ted Downing explained that he and Senator Karen Johnson drafted the initial audit bill in 2006. At the time, election officials from Pima and Maricopa counties (specifically Pima’s Brad Nelson and Maricopa’s Karen Osborne) cautioned that including county contests would result in the bill’s failure. To overcome this obstacle, the sponsors decided to exclude county races, ensuring the bill’s passage. This compromise was enacted into law—yet it excluded the contests most susceptible to insider manipulation, which were then exploited by the growth lobby.

As a result:

• Federal races are audited

• Statewide races are audited

• County races are not audited at all

Hand-count audits under A.R.S. §16-602 now provide the appearance of oversight without the capacity to detect problems.

Real verification requires ballot images, CVRs, and precinct-level reconstruction—the foundation of Auditable Ballot Examination (ABE).

Video: Former Representative Ted Downing explains how and why county races were excluded from Arizona’s audit law.