Auditing Elections Equipment In Maricopa County 

The Maricopa County Board of Supervisors voted unanimously on January 27, 2021 to authorize a forensic audit of ballot tabulation equipment used in the 2020 elections. The audit is the culmination of a year-long effort by Maricopa County to ensure the accuracy of the federally- and state-certified hardware and software used to count votes.


Learn More About the Vote >
Find 2020 Election Security & Accuracy Information>
View the 2020 Audit Timeline >

Audit Overview   

150th logo
The multi-layered forensic audit is comprised of three separate audits performed by two United States Elections Assistance Commission Voting System Testing Laboratories and a Certified Public Accounting Firm.

The audit of the tabulation equipment’s software and hardware was performing by two independent Voting System Testing Laboratories: Pro V&V and SLI Compliance. The firms were hired to:

  • Analyze election equipment software and hardware’s hacking vulnerability
  • Verify that no malicious malware was installed
  • Test that tabulators were not sending or receiving information over the internet
  • Confirm that no vote switching occurred


As an added measure of transparency, the Board hired a reputable Certified Public Accounting firm, Berry Dunn, to review county contracts with Dominion Voting Services and verify that the county leased the tabulation equipment according to state and county procurement regulations. That portion of the audit is expected to be completed by the end of the March.


Audit Results   

Maricopa County’s election equipment and software passed all tests performed by two independent firms hired to conduct the forensic audit, according to reports by two federally certified Voting System Testing Laboratories.

Read the Summary of the Audit Findings >
SLI Compliance Audit Report >
Pro V&V Audit Report >

1) Source Code Tests 

Source-Code

These tests were designed to verify if the same federally and state certified source code was installed on the equipment. Auditors conducted a full forensic clone of the drive which is a "bit-by-bit” copy. This allowed them to review deleted files, deleted file fragments, and hidden data that may be found in slack and unallocated space Source code extractions tested:

  • 20% of Election Day Tabulators (each firm tested 10%)   NO ISSUES 
  • 100% of Central Count Tabulators   NO ISSUES 
  • 100% of Election Management System work stations and servers
    NO ISSUES
  • 40% of Adjudication Stations (each firm tested 20%)   NO ISSUES

Results:

Pro V&V and SLI Compliance found that all software and equipment inspected was using certified software and was not modified.

2) Malicious Software and Hardware Tests

Hardware-Test2

These tests looked for evidence of equipment tampering or software hacking. Auditors ran tests for viruses, malicious software, and foreign equipment or code that should not be installed on the tabulators or system on the following equipment:

  • 20% of Election Day Tabulators (each firm tested 10%)   NO ISSUES 
  • 100% of Central Count Tabulators   NO ISSUES 
  • 100% of Election Management System work stations and servers
    NO ISSUES 
  • 40% of Adjudication Stations (each firm tested 20%)   NO ISSUES 

Results:

Pro V&V and SLI Compliance found no instances of malicious software or hardware installed on the tabulators or system.

3) Network and Internet Connectivity Tests

Network-and-Connectivity-Tests2These tests looked for evidence of the tabulations system ability to connect to the internet and if the tabulators and/or system was transmitting information outside the closed air gapped system within the county tabulation center or while being delivered, returned, or used at a vote center.

  • SLI reviewed system registry, system files, installed programs, download history, audit logs, USB/Ethernet history logs.   NO ISSUES 
  • Pro V&V ran server commands to test for connectivity.   NO ISSUES 
  • Both firms physically inspected and traced the wiring of the air-gapped system between the server, tabulators, adjudication stations, and Election Management System workstations. NO ISSUES 

Results:

Pro V&V and SLI Compliance found no evidence of internet connectivity.

4) Accuracy Tests (Only in Pro V&V Scope of Work)

Tabulation-Accuracy2These tests checked that the tabulation equipment and software correctly captures, stores, consolidates, and reports the specific ballot selections, and absence of selections, for each ballot position.


  • Using test ballots from the November General Election, Pro V&V examined the accuracy of more than 1.5 million ballot positions.
    NO ISSUES 


Results:

Pro V&V found no evidence of vote switching and concluded that the equipment tabulated and adjudicated ballots accurately.



Myth Busters



Both Pro V&V and SLI Compliance are certified Voting System Test Laboratories and their certification is in good standing with the U.S. Elections Assistance Commission

The county sought out the best, most qualified firms to perform this audit. A certified Voting System Test Laboratory is the only firm with the combined skill sets and knowledge to identify vulnerabilities and to verify the systems are configured according to federal certification standards.

Using a Voting System Test Laboratory is essential, as the county asked auditors to review and analyze at the tabulation source code to perform a true assessment and identify vulnerabilities. Since the Department of Homeland Security designated tabulation equipment as critical Election Infrastructure, if the source code is exposed and posted online, that could jeopardize the integrity of voting systems across the country.

With four jurisdictional elections in 2021, including one on March 9, the county must also ensure its tabulation equipment remains certified.



The test themselves were very thorough. In their hardware tests, auditors investigated the tabulation equipment by taking it apart piece-by-piece to the core components. They traced these core components back to the certified design build, ensuring there was no foreign hardware installed.

In their software tests, auditors took a bit-by-bit clone of the devices and every event log and audit log of the system.

They used five different antivirus and malware forensic examining tools to scan the bit-by-bit clone and search for any malicious software or viruses that would have made the system vulnerable.

They used five different antivirus and malware forensic examining tools to scan the bit-by-bit clone and search for any malicious software or viruses that would have made the system vulnerable.

These scans would have found if any devices connected to the internet or if an external USB drive was plugged into the election equipment. This testing was so thorough that when the Elections Department compiled the data and logs for the Senate’s subpoena, auditors were able to see exactly the steps taken.

If a malicious piece of hardware or software was installed, these tests would have found it.



All election specific data is stored on memory cards sealed in the vault or on our backup server. In addition, when the auditors performed their work, they made clones of our hard drives using write protection hardware or used brand new USB drives to gather the data. This would prevent any data corruption and maintained the air gapped integrity of the ballot tabulation center.

All the election equipment data from the General Election is stored in a separate file from the Goodyear election and cannot be overwritten.



The firms have attested they have not hired Dominion or any other former voting systems employees. Furthermore, the U.S. Election Assistance Commission provides hiring restrictions for certified Voting System Test Laboratories. If a firm were to hire a former voting system employee, that employee would be barred from development and testing of equipment of their former employer for at least three years.

Audit Firms Staff Question

Download Graphic >






During the federal certification process, Maricopa County’s election equipment’s software and hardware went through a robust set of security tests designed to ensure the equipment did not allow for vote manipulation, external or backdoor access or any other hackable workaround.

A Voting System Test Laboratory would not recommend a voting system with security weaknesses. Federal security standards include:

  • Protection of the critical elements of the voting system
  • Establishing and maintaining controls to minimize errors
  • Protection from intentional manipulation, fraud and malicious mischief; fraudulent or erroneous changes to the voting system
  • Protecting the secrecy in the voting process

A match of the software and hardware confirms that the system went through rigorous security testing.



The ballots are currently sealed. There is no provision under Arizona law that allows for a “recount” or “audit” of the election results after they were certified. The ballots cannot be accessed without a court order, this includes digital copies of the ballots.



The County provides a public wireless network so customers, including candidates and voters, can access the internet while doing county business. It is not unique to the Elections Department. The tabulation equipment is an air-gapped system, meaning it does not have access to any wireless network, including the public network. The two Voting System Test Laboratories confirmed that the tabulation systems was a secure closed network.















For more forensic audit photos, visit the Audit Timeline Multi-media Page.