Keeping Big Data from Big Brother

Published: May 23, 2013

As Big Data’s potential unfolds, how will privacy concerns be allayed?

Photo: Bob Perachio - RDPphotography.comPhoto: Bob Perachio - RDPphotography.com

Revolutions overturn existing power structures.

That reminder, given to CIOs by MIT Sloan professor Erik Brynjolfsson at the MIT Sloan CIO Symposium yesterday, made an implicit point clear: we are in the midst of a Big Data revolution. There will be winners, there will be losers, and it will be messy along the way.

Big Data is not just a revolution in technology, but in management, Brynjolfsson said. Better data means better ways to manage.

But as Big Data—large, complex sets of information that have rapidly expanded in size and scope with the widespread use of the Internet and mobile devices—becomes ubiquitous in business and government, the messiest by-product may be the threat to privacy, data ownership, and in some cases, civil rights.

Brynjolfsson’s research offers a positive outlook for business leaders and CIOs. Data-driven decision-makers, Brynjolfsson said, see 5 percent higher productivity, 6 percent greater profit, and 50 percent higher market value from information technology.

The possibilities are enormous. For example, MIT Sloan professor Dimitris Bertsimas shared his work using massive data sets to optimize clinical trials of cancer treatment. The results, he showed, suggest that data can point to a more optimal treatments than the average oncologist. MIT Sloan finance professor Andrew Lo predicted Big Data will lead to a more robust global financial system.

But the panel—while optimistic about what Big Data will eventually do for the world, for business, and for society—was not naïve to concerns about privacy and individual rights.

It will be up to the leaders of the Big Data revolution, panel members said, to navigate those issues so the revolution’s full effect can be felt. Or, as Lo put it, “to prevent Big Data from becoming Big Brother.”

Lo said cryptography will be a key element of convincing people that personal data can be used without violating privacy.

“That’s about the only way we’re going to ensure privacy on a level people are comfortable with,” Lo said. In a later panel about health care IT, panelists discussed ways to remove identifying information from large data sets of personal health records, allowing it to be used in the aggregate by hospitals and medical researchers.

MIT professor Alex “Sandy” Pentland, director of the MIT Media Lab Entrepreneurship Program, showed how cell phone data in Ivory Coast was used to decrease commute times by 10 percent, thanks to tracking riders’ origins and destinations. But collecting data also provided Ivory Coast with fresh census results, a data set with serious implications in a country which only two years ago fought a civil war based on claims of discrimination.

In law enforcement, programs like the New York City Police Department’s CompStat process—which uses data to track and target crime—are considered widely successful in the law enforcement community and are replicated in departments across the country. But the same principles are at the center of the department’s controversial Stop-and-Frisk program, which is accused of targeting black and Latino residents.

Asked Brynjolfsson, by all means a futurist with a positive outlook: “What kind of society are we going to be when we start thinking we can predict what people are going to do before they do it?”

Still, Pentland remained optimistic that education and positive results from Big Data projects will help alleviate privacy and data ownership concerns, citing Lo’s emphasis on the use of cryptography to enhance privacy while allowing data to be used to its full potential.

“The real challenge here is making the data available,” Pentland said. “We have to make it so that people are willing to … share data.”