by Jayson Hawkins

When George Orwell imagined the dystopian society of 1984, humanity simply did not possess the technology to bring his vision of the omnipresent surveillance state to life, even if there was no shortage of governments that wanted to do just that. The simple fact is that collecting, sorting, and analyzing intelligence data has always been a clunky, labor intensive, and time consuming affair, and the inherent inefficiency of large-scale spying operations has made it next to impossible for even the most paranoid police states to create a real-world Big Brother. Technology, however, inevitably advances, and within the last few decades, a combination of technical advances, cultural trends, and public ambivalence has brought the modern world closer than ever to the realization of security forces that see, hear, and know all.

Throughout the history of civilization, the agents of the state, both police and military, have had to act on incomplete information. Spies of all sorts are, of course, as old as time, but even as late as the end of the twentieth century, the constant complaint heard from the heads of police agencies was that they needed more and better intelligence gathering.

In the post-9/11 world, however, the complaint has changed. Police and intelligence officials now complain that they are, as one Chicago police official put it, “data rich but information poor.” The problem had become surveillance overload. With so much intelligence coming in from the massive spying arsenal of security cameras, license tag readers, and the unapologetic surveillance of social media, police agencies saw everything but knew very little.

The problem is the process known in intelligence-gathering circles as “fusion.” Fusion takes the data from various streams, gauges its value, sorts it, and finds the patterns that make the intelligence useful. This has historically been an intensely human process.

Typically, each human analyst is responsible for one data stream.

Teams of analysts then compare their data to search for patterns. Experiments with computerized fusion had never been able to create a system that could handle two or three data streams; modern police intelligence units contend with hundreds every day.

Enter DARPA (Defense Advanced Research Projects Agency), the research and development outfit at the U.S. Department of Defense. Military officials were suffering from the same data­overload as police agencies. Dan Kaufman, DARPA’s director of information innovation, reached out in 2010 to Ben Cutler, an engineer from the private sector who was building a new operating system at Microsoft. Together, they worked up a proposal for an innovative platform called Insight that could find ways to correlate “hard” data like GPS coordinates with “soft” data like informant reports, all while contextualizing data with past intelligence. To sync up the different data streams, gauge their quality, and account for gaps and ambiguity would require complex algorithms and a platform capable of crunching thousands of gigabytes at a time. Then, most importantly, the system would have to incorporate a type of predictive artificial intelligence, in essence learning to act like a human analyst.

The fact that most of these requirements seemed like science fiction fantasies did not deter Kaufman or Cutler. By 2013, Insight was close enough to being operational that the program was transferred from DARPA to the Army. Two years later, the Army was running massive Insight-driven battlefield simulations at Fort Irwin in California. At that point, the public profile of Insight disappears, but the technology has reemerged in a troubling new form.

In the fall of 2019, Arthur Holland Michel, a senior fellow at the Carnegie Institute for Ethics in International Affairs, was given a tour of several fusion centers being developed by private firms for use by police agencies in the U.S. and Europe.

The first tour was at the Paris office of a company called Genetec, where the head of the firm’s public safety division, Giovanni Gaccione, showed off a new flagship fusion product called Citigraf.

The demonstration showed Citigraf at work in Chicago, the first major city to deploy the program. A huge screen showed a map of Chicago’s East Side. Around the edges were thumbnail video streams from surveillance cameras and pop-up icons corresponding to unfolding emergencies. Each pop-up had a button marked “investigate.”

Holland Michel clicked on an icon marking a 911 call reporting an assault, and immediately a series of leads appeared including people in the neighborhood previously arrested for assault, a list of parolees living in the area, and the license plates of vehicles that had recently left the area, he wrote in a Wired magazine article published February 4, 2021. The longer the program ran, the more information it displayed, eventually gathering what Holland Michel called “more than enough information ... for an officer to respond to that original 911 call with a nearly telepathic sense of what had just unfolded.”

Genetec also demonstrated a program called Valcri. Originally designed to target sex-trafficking networks, Valcri has been upgraded with more sophisticated algorithms to do the more nuanced analysis associated with long-term detective work. Gaccione proudly claimed that Valcri would do investigative work that would take a human team weeks in less than a day.

Genetec is not alone. Holland Michel got extraordinary access to a fusion center in New York designed for the city by Microsoft where officers demonstrated the ease with which they could enter a name and get what “felt like more personal information than perhaps a curious officer, had any right to know without a court order.” The European Union is developing its own system, called ROXANNE (or, Real time network, text, and speaker analytics for combating organized crime), which is designed to unmask criminal networks and their members.Cisco, Palantir, and Motorola all also sell next generation fusion systems globally.

These new systems at the most basic level are built on the Insight model developed at DARPA. It is unclear if any of the algorithms developed for Insight are being utilized in these new platforms, but what is clear is that technology developed for conventional battlefield awareness and counter-terrorism operations is now in the hands of local police offices and being used to monitor neighborhoods across the nation.

The growth of the surveillance state is not the result of technical advances alone. Public acquiescence to the ubiquity of surveillance cameras and the wholesale embrace of privacy-crushing digital technology has coupled with a lack of sustained outrage as each new government-spies-on-citizens scandal fades from memory to embolden agencies to deploy ever more invasive surveillance systems. Without sustained pushback from a citizenry that is jealous of its privacy, Orwell’s nightmare of an omniscient state security apparatus may become a reality sooner than anyone expected. 

Sources: zerohedge.com, wired.com, theburkean.ie

Subscribe today

Already a subscriber? Login