Will Future Virtual Intelligence & Precrime Predictions Kill Privacy?

If we are all connected by 6 degrees of separation, how long before you comment, tweet, or link to a post that is part of precrime intelligence?

If crime could be stopped before it started, by use of virtual intelligence "future" searches, what might be the cost to privacy and online anonymity? When U.S. Intelligence communities have a strong mutual interest in the same projects as software companies whose products we use daily like Google or Microsoft, it raises more questions than it answers. If you recall Minority Report, then you probably remember that guilt by precrime predictions didn't work out so well.

Google Ventures and In-Q-Tel, the investment arm of the CIA, have provided funding to a company that trawls over half a million websites, Twitter feeds, YouTube, and blog posts, looking for connections between people, groups, and events. The company, Recorded Future, offers a Temporal Analytics Engine for predictive analysis, allowing people to "visualize the future, past, or present."

The search capabilities are not new, but the government's interest in the 'invisible links' between people is what concerns me. Are we on the edge of future virtual surveillance which will predict crimes that an individual will commit?

First off, I'd like what qualifies as public domain to be less fuzzy and more defined. If information is scraped and analyzed from public domains, then what exactly do Google and Microsoft consider public domains? If you save a draft in an email, but don't send it, is it public domain if it's merely stored on software companies' servers? The intelligence community has a history of warrantless seizure of emails without notifying the account holder. In the past, the EFF has fought this battle. "Government investigators maintain that because the Yahoo! email has been accessed by the user, it is no longer in 'electronic storage' under the Stored Communications Act (SCA) and therefore does not require a warrant."

If stored communications lead to predicting crime, then it seems less like data mining and more like monitoring and censorship. But what if predictive analysis does not include stored drafts or documents? Could future virtual intelligence still "tag" us by virtue of association?

Six degrees of separation suggests that any two people can be made to connect in six steps or fewer. According to Wikipedia, the same theory of six intermediaries applies to email and chat. "A 2007 study by Jure Leskovec and Eric Horvitz examined a data set of instant messages composed of 30 billion conversations among 240 million people. They found the average path length among Microsoft Messenger users to be 6.6." What if a friend of a friend has "hot keywords" in an email or an instant message? Are you then somehow also associated to a possible precrime by virtue of six degrees of separation?

Although Google and the CIA are interested in predictive analysis, Recorded Future is far from the first to monitor the web  for trends and apply the scouring of information to future predictions. Here are a few others:

The Web Bot Project, created in 1997, sends software spiders to search the Internet and make predictions. "The technology is claimed to be able to examine the collective unconscious and be able to predict catastrophic events 60 to 90 days in advance." Web Bot claims to have correctly predicted events such as the 9/11 attacks and the BP oil spill. It has also had its share of predictions that missed the mark.

IBM's Jeff Jonas helped develop Systems Research & Development (SRD), a facial recognition technology used for surveillance intelligence. According to IBM, "Following an investment in 2001 by In-Q-Tel, the venture capital arm of the CIA, SRD began playing a role in America's national security and counterterrorism mission. One such contribution includes a unique analysis of the connections between the 9/11 terrorists. This 'link analysis' is so unique that it is taught in universities and has been the widely cited by think tanks and the media."

Psydex provides real time search, data mining, and predictive analytics used by DoD and U.S. Intelligence agencies to sort through the sea of online chatter. Psydex case studies note,  "A government agency has a team of individuals responsible for monitoring open source information, leveraging the eyes and ears of the world to detect potential civil unrest, disasters, acts of terror, and other major events in near real time. Open Source Intelligence (OSINT) has become more relevant as social media and blogs have created a surge in publicly disseminated information."

Big Brother using link or predictive analysis is not new. It's also not the first time U.S. Intelligence communities joined forces with giant software companies in the name of security. NSA assisted Apple and Linux in the development of security and operating systems. Microsoft credited the NSA for helping in the development of Windows operating systems. According to NSA about Windows Vista, "NSA personnel worked hand-in-hand with Microsoft" on security settings "desirable for sensitive and traditional high security network environments." Microsoft and the NSA also collaborated on Windows XP and Windows Server 2003 security guidance.

If the Microsoft and NSA collaboration reassured you about Windows security, then perhaps you may like the idea of the CIA and Google funding the same "future" of web monitoring startup? Personally, I don't find it reassuring. I don't much like what happens to privacy in the name of security. If we are all connected by six degrees of separation, how long before you comment, tweet, or link to a post that is part of precrime intelligence?

Like this? Check out these other posts:

Follow me and all the other Microsoft Subnet bloggers on Twitter @microsoftsubnet

Copyright © 2010 IDG Communications, Inc.

Get the best of CSO ... delivered. Sign up for our FREE email newsletters!