Artificial Intelligence Can Hunt Down Missile Sites in China Hundreds of Times Faster Than Humans


Intelligence agencies have a limited number of trained human analysts looking for undeclared nuclear facilities, or secret military sites, hidden among terabytes of satellite images. But the same sort of deep learning artificial intelligence that enables Google and Facebook to automatically filter images of human faces and cats could also prove invaluable in the world of spy versus spy. An early example: US researchers have trained deep learning algorithms to identify Chinese surface-to-air missile sites--hundreds of times faster than their human counterparts. The deep learning algorithms proved capable of helping people with no prior imagery analysis experience find surface-to-air missile sites scattered across nearly 90,000 square kilometers of southeastern China. Such AI based on neural networks--layers of artificial neuron capable of filtering and learning from huge amounts of data--matched the overall 90 percent accuracy of expert human imagery analysts in locating the missile sites.

Amazon announces AWS Secret Region for intelligence agencies


Amazon Web Services (AWS) has announced setting up a "secret" datacentre region targeted towards the US intelligence community and other government agencies working with secret-level datasets. AWS Secret Region is able to host software and data that are classified at the "secret" level, making it applicable to intelligence agencies that typically deal with sensitive information. Secret Region is an extension of the $600 million AWS-Central Intelligence Agency arrangement that led to the creation of Top Secret Region in 2014 specifically for the US intelligence community. The new region is immediately available to US intelligence agencies through their existing commercial cloud services contract with AWS and will meet certain government standards. But it will also be available to other types of government customers with sufficient secret-level network access and their own "contract vehicles".

At a Glance – Adversarial Attacks - Disruption Hub


No machine learning algorithm is perfect. Whilst the margin of error might be tiny, any computer which uses such algorithms sometimes makes mistakes. Earlier this month, research conducted by a team of students from MIT showed that Google's neural network could be tricked into misidentifying a 3D printed turtle as a gun. The group used a hacking technique known as an adversarial attack, altering the image that the software received. In other words, an adversarial attack is a smokescreen for computers.

Threat intelligence platform adds analyst assessments to machine learning


Companies are increasingly turning to AI and machine learning solutions to combat cyber threats, but sometimes there is no substitute for the insight that comes with human analysis. Threat intelligence specialist Recorded Future recognizes this and is expanding its platform to give security operations centers access to analyst-originated intelligence to offer relevant expert insights and analysis needed for operational improvements and targeted risk reduction. "To effectively combat the risks of cyber attacks, defenders need intelligence from the widest range of sources in real-time," says Dr Christopher Ahlberg, CEO and co-founder at Recorded Future. "The direct access to analyst insights combined with our open, closed and technical threat intelligence sources provides our customers with the most powerful source of advantage against their adversaries. The breadth of threat intelligence sources we arm customers with is unmatched and puts organizations in the best position possible to defend against threats."

Data-Protection Efforts Must Prepare for New Forms of Attack


Organizations already have plenty to worry about in terms of data protection, but a new type of cyberattack could prove much more damaging and harder to remediate. A destruction of service (DeOS) attack has the potential to destroy the data backups and safety nets organizations rely on to restore their systems and data following an attack, according to Cisco. DeOS attacks are a more dangerous version of distributed denial of service (DDoS), which employs botnets to overload the target organization's servers with traffic until they can no longer handle the extra load. DDoS attacks last hours or days, after which a company can resume normal operations. This is one of the many new security risks that are emerging with the Internet of Things (IoT).

AI and the Future of Network Security


According to a Cisco whitepaper examining the rapid expansion of the internet of things, there will be more than 50 billion internet-connected devices in the world by 2020, representing a hundred-fold increase since 2003. This proliferation of connected devices is making our everyday lives and work easier, but the convenience comes with a number of risks. Research from PwC indicates that the number of worldwide cybersecurity incidents rose by 38% in 2015, the largest increase in any of the 12 years the firm has conducted its Global State of Information Security Survey. Due in large part to the ever-increasing number of connected devices in the workplace, the average North American enterprise is inundated by alerts from its cybersecurity systems; 2014 research estimated the number at 10,000. The most active enterprise networks often receive an astounding 150,000 alerts per day.

US launches Libya drone strike as Africa operations appear to ramp up

FOX News

The Libyan National Army has been battling ISIS in the cities of Sirte and Benghazi. The U.S. military has launched airstrikes this month in Yemen, Somalia, Iraq, Syria, Afghanistan and Friday, for the first time since September, in Libya. According to a defense official, the drone strike in the desert of central Libya Friday killed "several" ISIS militants in a sign the Pentagon may be ramping up pressure on terror groups in Africa. The most recent strike comes a year after the military launched nearly 500 airstrikes against ISIS in the coastal city of Sirte, located halfway between Tripoli and Benghazi. The September strike killed 17 ISIS fighters.

Back-Flipping Robot Is A Giant Leap For Robot Kind


MIT's Atlas robot, nicknamed Helios, completes the driving task at the June 2015 DARPA Robotics Challenge Finals. Helios is a second-generation Atlas, developed for DARPA by Boston Dynamics. MIT's Atlas robot, nicknamed Helios, completes the driving task at the June 2015 DARPA Robotics Challenge Finals. Helios is a second-generation Atlas, developed for DARPA by Boston Dynamics. Thursday, Nov. 16, 2017 was a day filled with news.

Achieving Accurate, Reliable AI Trajectory Magazine


What will happen to a person's artificial intelligence (AI) when they retire? When a prospective employee interviews for a job, will his or her AI be questioned alongside them? Will companies hire AI straight from a factory, or will the system undergo a sort of apprenticeship before being put to work? More importantly--and more realistic in the near-term--what will be the line at which machines are not reliable enough or morally appropriate to use and humans take over? These, along with many more immediate questions, are among the topics USGIF's Machine Learning & Artificial Intelligence Working Group seeks to generate discussion around.

3 Questions: Lisa Parks on drones, warfare, and the media

MIT News

Drones have become a common part of warfare -- but their use remains a subject of public contention. Lisa Parks, a professor in MIT's program in Comparative Media Studies/Writing and director of its Global Media Technologies and Cultures Lab, has spent extensive time analyzing this public debate. Now, she has co-edited a new volume examining the subject, while contributing a piece to it herself. The book, "Life in the Age of Drone Warfare," has just been published by Duke University Press. MIT News talked with Parks this week about the impact and public perception of drones.