Computer expert still in Maelstrom Officials say voting machines secure

10/06/04
By Linda Strowbridge Owings Mills Times

He has alternately been called an enemy of democracy and a high-tech champion of civil rights.

Aviel Rubin, a soft-spoken computer scientist who lives in Owings Mills, has been a flash point in the national fight over electronic voting ever since he reported that Maryland's new voting machines might be vulnerable to tampering.

Although his claims have been hotly denied by the machines' manufacturer and Maryland elections officials, Rubin and other independent researchers continue to believe Diebold Elections Systems' equipment contains security flaws that could put the results of the Nov. 2 vote in doubt.

Rubin is working with members of Congress to pass legislation requiring tighter security and a paper audit trail from all electronic voting machines.

"This issue has consumed my life," said Rubin, a computer science professor at Johns Hopkins University and technical director of the Information Security Institute on Hopkins' Homewood campus. "But it just seems too important to let go."

Rubin, an accomplished but low-key computer security expert, was catapulted onto the front pages of The Washington Post, The New York Times and other national news outlets in July 2003, when he released a report on Diebold's electronic voting machines.

Rubin and fellow researchers concluded, "The voting system is far below even the most minimal security standards."

They reported unauthorized people could conceivably access the tallies on individual machines and disable further voting, use homemade smart cards to vote multiple times or tap into the system through telephone or network connections and view or even alter votes.

Coincidentally, Rubin's report was released to the media just one day before Maryland's Board of Elections announced a contract for $55 million worth of Diebold machines.

"It was a total nightmare for Maryland elections officials," Rubin said. "They were so proud they had made this move to modern technology, and then we came out of left field."

Diebold immediately denounced the study as inaccurate, naive and shallow. The company claimed Rubin had analyzed only a portion of the system's code and inappropriately tested it on a computer rather than a voting machine. Rubin said the company's lawyers sent a letter to him urging him to cease and desist all discussions of their voting machines or face legal action.

The state elections board said Rubin had failed to factor in the precautions officials take to safeguard voting machines. Insisting that the machines were secure, Board of Elections Administrator Linda Lamone said Rubin's study did a "great disservice to democracy."

Rubin was suddenly swept up in a frenetic controversy. Presidential hopefuls, including Howard Dean, Dennis Kucinich and Carol Moseley Braun contacted him, as did members of the House of Representatives. He testified at congressional hearings, spoke to computer security conferences and participated in innumerable news stories and documentaries.

A new and untenured professor at Johns Hopkins, Rubin uncomfortably realized he had become a public and controversial figure in the staid halls of academia.

"Since I was a new faculty member, I was invited to the new-faculty welcome dinner at the president's house," Rubin recalled. When President William Brody approached, "I think the first words he said were, 'Oh, you're the troublemaker.'"

"But every subsequent study verified, substantiated and reinforced the conclusions that Avi came to," said Gerald Masson, director of Hopkins' Information Security Institute.

A report by RABA Technologies, commissioned by the state legislature, and another by SAIC Corp., commissioned by Gov. Robert Ehrlich, also warned that the Diebold system contained vulnerabilities that could be exploited by people seeking to change election results.

"I don't think I would want to try to fix that system," Rubin said. "This thing is Swiss cheese. It's full of bugs and, trying to fix it, you're just going to make it worse. Imagine a building whose foundation is made from materials that disintegrate. At some point, it is easier to demolish it and build a new structure than to try to continually patch it."

Any electronic system should produce a printout so voters can ensure their votes are properly recorded, and elections officials can conduct proper recounts, Rubin said.

It's a point he has pressed before.

As author of four books on computer security, Rubin prompted the government of Costa Rica to abort plans to adopt electronic voting in 1997, when he detailed the security flaws in that system.

His analysis of security problems with SERV _ a system to enable military personnel and civilians overseas to vote via the Internet _ prompted Defense Undersecretary Paul Wolfowitz to scrap the system even though the Defense Department had spent $22 million on SERV.

In July 2003, Rubin hoped his Diebold report would spark similar interactions with Maryland's elections officials. But the invitation never came.

Jacqueline McDaniel, director of the Baltimore County Board of Elections, said she believes the voting system is safe. A team of 27 county officials has been conducting tests on machines that will be used in the November vote and found them to be secure and reliable, she said.

"The more you work on them, the more you see how safe they are," McDaniel said.

"The likelihood of someone being able to tamper with the machines is nonexistent," said David Bear, a spokesman for Diebold Elections Systems. He said the company made some security modifications as a result of the independent studies.

He said the voting machines will reduce the number of Americans who are disenfranchised because elections systems don't accurately record their votes. In the vote that elected Arnold Schwarzenegger as governor of California, elections machinery failed to capture the vote of nearly seven percent of people using punch cards, but missed only 0.73 percent of votes cast by people at touch screen voting machines.

But Rubin, who worked as an election judge in the Maryland primary, said his experience with the Diebold machines gave him "more of an appreciation of how insecure the system is."

"I am very disappointed and worried," he said.

Despite security precautions and accuracy testing, Rubin fears the system will not be able to track errors or tampering in the Nov. 2 vote.

"We could end up having an election where, if it's close and the outcome comes into question, we can't do recounts. We could have a population that believes the election wasn't done fairly," he said.

"In 2000, people complained about hanging chads. But a hanging chad is better than no ballot. Now we are in the situation of where we will have no ballots."