Centre for Internet & Society
Hacking without borders: The future of artificial intelligence and surveillance

Source: TheoJunior on flickr

In this post, Maria Xynou looks at some of DARPA´s artificial intelligence surveillance technologies in regards to the right to privacy and their potential future use in India.


This research was undertaken as part of the 'SAFEGUARDS' project that CIS is undertaking with Privacy International and IDRC.


Robots or computer systems controlling our thoughts is way beyond anything I have seen in science fiction; yet something of the kind may be a reality in the future. The US Defence Advanced Research Projects Agency (DARPA) is currently funding several artificial intelligence projects which could potentially equip governments with the most powerful weapon possible: mind control.

Combat Zones That See (CTS)

Source: swanksalot on flickr

Ten years ago DARPA started funding the Combat Zones That See (CTS) project, which aims to ´track everything that moves´ within a city through a massive network of surveillance cameras linked to a centralized computer system. Groundbreaking artificial intelligence software is being used in the project to identify and track all movement within cities, which constitutes Big Brother as a reality. The computer software supporting the CTS is capable of automatically identifying vehicles and provides instant alerts after detecting a vehicle with a license plate on a watch list. The software is also able to analyze the video footage and to distinguish ´normal´ from ´abnormal´ behavior, as well as to discover links between ´places, subjects and times of activity´ and to identify patterns. With the use of this software, the CTS constitute the world´s first multi-camera surveillance system which is capable of automatically analyzing video footage.

Although the CTS project was initially intended to be used for solely military purposes, its use for civil purposes, such as combating crime, remains a possibility. In 2003 DARPA stated that 40 million surveillance cameras were already in use around the world by law enforcement agencies to combat crime and terrorism, with 300 million expected by 2005. Police in the U.S. have stated that buying new technology which may potentially aid their work is an integral part of the 9/11 mentality. Considering the fact that literally millions of CCTV cameras are installed by law enforcement agencies around the world and that DARPA has developed the software that has the capability of automatically analyzing data gathered by CCTV cameras, it is very possible that law enforcement agencies are participating in the CTS network.

However if such a project was used for non-military level purposes, it could raise concerns in regards to data protection, privacy and human rights. As a massive network of surveillance cameras, the CTS ultimately could enable the sharing of footage between private parties and law enforcement agencies without individuals´ knowledge or consent. Databases around the world could be potentially linked to each other and it remains unclear what laws would regulate the access, use and retention of such databases by law enforcement agencies of multiple countries. Furthermore, there is no universal definition for ´normal´ and ´abnormal´ behaviour, thus if the software is used for its original purpose, to distinguish between “abnormal” and “normal” behaviour, and used beyond military purposes, then there is a potential for abuse, as the criteria for being monitored, and possibly arrested, would not be clearly set out.

Mind´s Eye

Source: watchingfrogsboil on flickr

A camera today which is only capable of recording visual footage appears futile in comparison to what DARPA´s creating: a thinking camera. The Mind´s Eye project was launched in the U.S. in early 2011 and is currently developing smart cameras endowed with ´visual intelligence´. This ultimately means that artificial intelligence surveillance cameras can not only record visual footage, but also automatically detect ´abnormal´ behavior, alert officials and analyze data in such a way that they are able to predict future human activities and situations.

Mainstream surveillance cameras already have visual-intelligence algorithms, but none of them are able to automatically analyze the data they collect. Data analysts are usually hired for analyzing the footage on a per instance basis, and only if a policeman detects ´something suspicious´ in the footage. Those days are over. General James Cartwright, the vice chairman of the Joint Chiefs of Staff, stated in an intelligence conference that “Star[ing] at Death TV for hours on end trying to find the single target or see something move is just a waste of manpower.” Today, the Mind´s Eye project is developing smart cameras equipped with artificial intelligence software capable of identifying operationally significant activity and predicting outcomes.

Mounting these smart cameras on drones is the initial plan; and while that would enable military operations, many ethical concerns have arisen in regards to whether such technologies should be used for ´civil purposes.´ Will law enforcement agencies in India be equipped with such cameras over the next years? If so, how will their use be regulated?

SyNAPSE

Source: A Health Blog on flickr

The Terminator could be more than just science fiction if current robots had artificial brains with similar form, function and architecture to the mammalian brain. DARPA is attempting this by funding HRL Laboratories, Hewlett-Packard and IBM Research to carry out this task through the Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) programme.  Is DARPA funding the creation of the Terminator? No. Such artificial brains would be used to build robots whose intelligence matches that of mice and cats...for now.

SyNAPSE is a programme which aims to develop electronic neuromorphic machine technology which scales to biological levels. It started in the U.S. in 2008 and is scheduled to run until around 2016, while having received $102.6 million in funding as of January 2013. The ultimate aim is to build an electronic microprocessor system that matches a mammalian brain in power consumption, function and size. As current programmable machines are limited by their computational capacity, which requires human-derived algorithms to describe and process information, SyNAPSE´s objective is to create biological neural systems which can autonomously process information in complex environments. Like the mammalian brain, SyNAPSE´s cognitive computers would be capable of automatically learning relevant and probabilistically stable features and associations, as well as of finding correlations, creating hypotheses and generally remembering and learning through experiences.

Although this original type of computational device could be beneficial to predict natural disasters and other threats to security based on its cognitive abilities, human rights questions arise if it were to be used in general for surveillance purposes. Imagine surveillance technologies with the capacity of a human brain. Imagine surveillance technologies capable of remembering your activity, analyzing it, correlating it to other facts and/or activities, and of predicting outcomes; and now imagine such technology used to spy on us. That might be a possibility in the future.

Such cognitive technology is still in an experimental phase and although it could be used to tackle threats to security, it could also potentially be used to monitor populations more efficiently. No such technology currently exists in India, but it could only be a matter of time before Indian law enforcement agencies start using such artificial intelligence surveillance technology to supposedly enhance our security and protect us.

Brain-Computer Interface (BCI)


Remember Orwell's ´Thought Police´? Was Orwell exaggerating just to get his point across? Well, the future appears to be much scarier than Orwell's vision depicted in 1984. Unlike the ´Thought Police´ which merely arrested individuals who openly expressed ideas or thoughts which contradicted the Party´s dogma, today, technologies are being developed which can literally read our thoughts.

Once again, DARPA appears to be funding one of the world´s most innovative projects: the Brain-Computer Interface (BCI). The human brain is far better at pattern matching than any computer, whilst computers have greater analytical speed than human brains. The BCI is an attempt to merge the two together, and to enable the human brain to control robotic devices and other machines. In particular, the BCI is comprised of a headset (an electroencephalograph - an EEG) with sensors that rest on the human scalp, as well as of software which processes brain activity. This enables the human brain to be linked to a computer and for an individual to control technologies without moving a finger, but by merely thinking of the action.

Ten years ago it was reported that the brains of rats and monkeys could control robot arms through the use of such technologies. A few years later brainstem implants were developed to tackle deafness. Today, brain-computer interface technologies are able to directly link the human brain to computers, thus enabling paralyzed people to conduct computer activity by merely thinking of the actions, as well as to control robotic limbs with their thoughts. BCIs appear to open up a new gateway for disabled persons, as all previously unthinkable actions, such as typing on a computer or browsing through websites, can now be undertaken by literally thinking about them, while using a BCI.

Brain-controlled robotic limbs could change the lives of disabled persons, but ethical concerns have arisen in regards to the BCI´s mind-reading ability.  If the brain can be used to control computers and other technologies, does that ultimately mean that computers can also be used to control the human brain?  Researchers from the University of Oxford and Geneva, and the University of California, Berkley, have created a custom programme that was specially designed with the sole purpose of finding out sensitive data, such as an individuals´ home location, credit card PIN and date of birth. Volunteers participated in this programme and it had up to 40% success in obtaining useful information. To extract such information, researchers rely on the P300 response, which is a very specific brainwave pattern that occurs when a human brain recognizes something that is meaningful, whether that is personal information, such as credit card details, or an enemy in a battlefield. According to DARPA:

´When a human wearing the EEG cap was introduced, the number of false alarms dropped to only five per hour, out of a total of 2,304 target events per hour, and a 91 percent successful target recognition rate was introduced.´

This constitutes the human brain as a new warfighting domain of the twenty-first century, as experiments have proven that the brain can control and maneuver quadcopter drones and other military technologies. Enhanced threat detection through BCI´s scan for P300 responses and the literal control of military operations through the brain, definitely appear to be changing the future of warfare. Along with this change, the possibility of manipulating a soldier´s BCI during conflict is real and could lead to absolute chaos and destruction.

Security expert, Barnaby Jack, of IOActive demonstrated the vulnerability of biotechnological systems, which raises concerns that BCI technologies may also potentially be vulnerable and expose an individual's´ brain to hacking, manipulation and control by third parties. If the brain can control computer systems and computer systems are able to detect and distinguish brain patterns, then this ultimately means that the human brain can potentially be controlled by computer software.

Will BCI be used in the future to interrogate terrorists and suspects? What would that mean for the future of our human rights? Can we have human rights if authorities can literally hack our brain in the name of national security? How can we be protected from abuse by those in power, if the most precious thing we have - our thoughts - can potentially be hacked? Human rights are essential because they protect us from those in power; but the privacy of our thoughts is even more important, because without it, we can have no human rights, no individuality.

Sure, the BCI is a very impressive technological accomplishment and can potentially improve the lives of millions. But it can also potentially destroy the most unique quality of human beings: their personal thoughts. Mind control is a vicious game to play and may constitute some of the scariest political novels as a comedy of the past. Nuclear weapons, bombs and all other powerful technologies seem childish compared to the BCI which can literally control our mind! Therefore strict regulations should be enacted which would restrict the use of BCI technologies to visually impaired or handicapped individuals.  Though these technologies currently are not being used in India, explicit laws on the use of artificial intelligence surveillance technologies should be enacted in India, to help ensure that they do not infringe upon the right to privacy and other human rights.

Apparently, anyone can buy Emotiv or Neurosky BCI online to mind control their computer with only $200-$300. If the use of BCI was imposed in a top-down manner, then maybe there would be some hope that people would oppose its use for surveillance purposes; but if the idea of mind control is being socially integrated...the future of privacy seems bleak.

The views and opinions expressed on this page are those of their individual authors. Unless the opposite is explicitly stated, or unless the opposite may be reasonably inferred, CIS does not subscribe to these views and opinions which belong to their individual authors. CIS does not accept any responsibility, legal or otherwise, for the views and opinions of these individual authors. For an official statement from CIS on a particular issue, please contact us directly.