Top CISA official pushes back against Elon Musk call to end electronic voting machines

An electronic ballot marking machine at a polling location during the "Super Tuesday" presidential primary in Hawthorne, California on March 5, 2024. A CISA official reaffirmed the security of U.S. voting infrastructure as the 2024 presidential election approaches.

An electronic ballot marking machine at a polling location during the "Super Tuesday" presidential primary in Hawthorne, California on March 5, 2024. A CISA official reaffirmed the security of U.S. voting infrastructure as the 2024 presidential election approaches. PATRICK T. FALLON/AFP via Getty Images

Allegations of voting machine flaws were a flash point in the 2020 presidential election.

A top official at the Cybersecurity and Infrastructure Security Agency on Tuesday reaffirmed the security of U.S. voting infrastructure, firing back at tech CEO Elon Musk’s post on the X social media platform calling for the elimination of electronic voting machines.

Musk in the June 15 post said the “risk of being hacked by humans or AI, while small, is still too high.” He quoted a separate post from the account of presidential candidate Robert F. Kennedy Jr. that linked an Associated Press story about a software issue in Dominion Voting Systems machines that led to incorrectly calculating vote totals in Puerto Rico’s June 2 primary.

The island is reviewing its contract with the company, according to the report, which notes that, while no one is contesting the primary results, the “machine-reported vote counts were lower than the paper ones in some cases, and some machines reversed certain totals or reported zero votes for some candidates.”

Asked about the Musk remarks, agency senior advisor Cait Conley said voting machine companies since 2016 have been doing “tremendous” work to make their systems secure against state-sponsored hackers and cybercriminals.

“When you look at things like the electronic systems that are used — especially when it comes to the votecasting or tabulating side — there are tremendous, tremendous efforts put in place to ensure their security through air gapping, though penetration testing … to again just ensure that these systems are secure,” she told an audience at Semafor event focused on cybersecurity policy.

“Going into 2024, we do have the highest degree of security for our election infrastructure than any other cycle,” she said. “The 2024 election cycle is more secure than any previous cycle.”

The use of AI tools to target voting systems isn’t the only concern on the table for the November presidential election. Voting equipment providers have been working to shore up defenses against hackers aiming to disable or change behaviors on voting machines. Touchscreen voting machines that glitched in the swing county of Northampton, Pennsylvania last year, for instance, raised concerns about voter confidence in the county ahead of what’s expected to be a high-profile election.

But the broader vulnerabilities of electronic voting systems have been called into question. A pair of reports released a year ago — one by a University of Michigan professor and another by the not-for-profit MITRE Corporation — dispute each other over how easily hackers could potentially hijack voting machines and change election outcomes. Machine providers have also walked a fine line between reporting vulnerabilities and keeping them private to prevent malicious actors from weaponizing them.

In an effort to shore up their defenses, voting machine vendors for at least the past year have already been making tweaks to their hardware while simultaneously running exercises that simulate attacks like ballot box stuffing or knocking electronic poll books offline.

Voting device flaws were a high-profile flash point for the 2020 election, as former president Donald Trump and conspiracy theorists lodged allegations that rigged machines led to President Joe Biden taking his role in the Oval Office. Trump has repeatedly called the 2020 results fraudulent and said voting machines in some states had malfunctioned. Those claims have been largely debunked.

Officials and researchers are focused on the convergence of AI-powered disinformation fueling physical election threats this year. Many experts in the cybersecurity field have expressed concerns with the level of mis- or disinformation that is flowing through the help of consumer-facing generative AI tools or related offerings available on the dark web.

The National Security Agency’s recently retired chief said in a private media briefing this month that he feels good about election integrity this year. “We’re going to know the threat better than they know themselves,” said Gen. Paul Nakasone. “We’re going to take all the information that we need, and we’re going to share it with both DHS and the FBI. And we’re going to take action when we see adversaries outside the United States trying to either influence or interfere in our election,” he added.