NSA searches for advanced data mining tech
Silicon Valley shopping spree
The National Security Agency (NSA) visited Silicon Valley this month on the hunt for private sector technology to beef up its already formidable snooping and signals intelligence portfolio. Data mining technologies to search for connections between seemingly unrelated snippets of information was top of the NSA's shopping list, according to venture capitalists who held meetings with agency officials.
The New York Times reports that the agency is hunting for technology that fits with its increased emphasis on scanning millions of ordinary Americans' phone calls and emails for 'suspicious' patterns. "The theory is that the automated tool that is conducting the search is not violating the law. [But] anytime a tool or a human is looking at the content of your communication, it invades your privacy," Mark D Rasch, former head of computer-crime investigations for the Justice Department and current SVP of computer security firm Solutionary, said.
According to the NYT, the NSA is seeking to extend data mining techniques (already widely used in the private sector by credit card companies and the like) by applying analysis tools used by law enforcement agencies in the search for potential terrorists. It lists a crime investigation spreadsheet called Analyst's Notebook, which allows investigators to view telephone and financial records of suspects, and visualisation tools from firms such as i2, as examples of the types of technologies the agency is interested in.
One security expert said the surveillance tools adopted as part of the Bush administration's anti-terrorist eavesdropping program are misguided. "In many respects, we're fighting the last intelligence war," John Arquilla, a professor of defence analysis at the Naval Postgraduate School in Monterey, California and a consultant on the Total Information Awareness project said. "We have not pursued data mining in the way we should."
Arquilla argued that the estimated $40bn a year spent on data mining programs would be better redirected in matching up information available from public source, such as "internet chat rooms used by Al Qaeda". ®