Google will not allow its artificial intelligence software to be used in weapons or unreasonable surveillance efforts under new standards for its business decisions in the nascent field, the Alphabet Inc unit said on Thursday. The restriction could help Google management defuse months of protest by thousands of employees against the company's work with the U.S. military to identify objects in drone video. Google instead will seek government contracts in areas such as cybersecurity, military recruitment and search and rescue, Chief Executive Sundar Pichai said in a blog post on Thursday. "We want to be clear that while we are not developing AI for use in weapons, we will continue our work with governments and the military in many other areas," he said. Breakthroughs in the cost and performance of advanced computers have carried AI from research labs into industries such as defense and health in the last couple of years.

The U.S. military has been looking to incorporate elements of artificial intelligence and machine learning into its drone program. Project Maven, as the effort is known, aims to provide some relief to military analysts who are part of the war against Islamic State. These analysts currently spend long hours staring at big screens reviewing video feeds from drones as part of the hunt for insurgents in places like Iraq and Afghanistan. The Pentagon is trying to develop algorithms that would sort through the material and alert analysts to important finds, according to Air Force Lieutenant General John N.T. 'Jack' Shanahan, director for defense intelligence for warfighting support. Military bosses say intelligence analysts are'overwhelmed' by the amount of video being recorded over the battlefield by drones with high resolution cameras'A lot of times these things are flying around(and)... there's nothing in the scene that's of interest,' he told Reuters.

It's never been clear whether robotics company Boston Dynamics is making killing machines, household helpers, or something else entirely. For nine years, the secretive firm -- which got its start with U.S. military funding -- has unnerved people around the world with YouTube videos of experimental robots resembling animal predators. In another, a small wheeled rover nicknamed SandFlea abruptly flings itself onto rooftops -- and back down again. A more recent effort features a slender dog-like robot that climbs stairs, holds its own in a tug-of-war with a human and opens a door to let another robot pass. Boston Dynamics has demonstrated little interest in elaborating on these glimpses into a possible future of fast, strong and sometimes intimidating robots. For months, the company and its parent, SoftBank, rebuffed numerous requests seeking information about its work.

SAN FRANCISCO (Reuters) - Google will not allow its artificial intelligence software to be used in weapons or unreasonable surveillance efforts under new standards for its business decisions in the nascent field, the Alphabet Inc (GOOGL.O) unit said on Thursday. The restriction could help Google management defuse months of protest by thousands of employees against the company's work with the U.S. military to identify objects in drone video. Google instead will seek government contracts in areas such as cybersecurity, military recruitment and search and rescue, Chief Executive Sundar Pichai said in a blog post bit.ly/2M8Pdkq on Thursday. "We want to be clear that while we are not developing AI for use in weapons, we will continue our work with governments and the military in many other areas," he said. Breakthroughs in the cost and performance of advanced computers have carried AI from research labs into industries such as defense and health in the last couple of years.

Cyber attacks are growing in frequency and severity. Over the past year alone we have witnessed massive data breaches that stole personal information of millions of people and wide-scale ransomware attacks that paralyzed critical infrastructure of several countries. Combating the rising cyber threat calls for a multi-pronged strategy, which includes predicting when these attacks will occur. The intuition driving our approach is this: during the planning and preparation stages, hackers leave digital traces of their activities on both the surface web and dark web in the form of discussions on platforms like hacker forums, social media, blogs and the like. These data provide predictive signals that allow anticipating cyber attacks. In this paper, we describe machine learning techniques based on deep neural networks and autoregressive time series models that leverage external signals from publicly available Web sources to forecast cyber attacks. Performance of our framework across ground truth data over real-world forecasting tasks shows that our methods yield a significant lift or increase of F1 for the top signals on predicted cyber attacks. Our results suggest that, when deployed, our system will be able to provide an effective line of defense against various types of targeted cyber attacks.

We propose a method to learn deep ReLUbased classifiers that are provably robust against normbounded adversarial perturbations on the training data. For previously unseen examples, the approach is guaranteed to detect all adversarial examples, though it may flag some non-adversarial examples as well. The basic idea is to consider a convex outer approximation of the set of activations reachable through a norm-bounded perturbation, and we develop a robust optimization procedure that minimizes the worst case loss over this outer region (via a linear program). Crucially, we show that the dual problem to this linear program can be represented itself as a deep network similar to the backpropagation network, leading to very efficient optimization approaches that produce guaranteed bounds on the robust loss. The end result is that by executing a few more forward and backward passes through a slightly modified version of the original network (though possibly with much larger batch sizes), we can learn a classifier that is provably robust to any norm-bounded adversarial attack. We illustrate the approach on a number of tasks to train classifiers with robust adversarial guarantees (e.g. for MNIST, we produce a convolutional classifier that provably has less than 5.8% test error for any adversarial attack with bounded l

We develop nested automatic differentiation (AD) algorithms for exact inference and learning in integer latent variable models. Recently, Winner, Sujono, and Sheldon showed how to reduce marginalization in a class of integer latent variable models to evaluating a probability generating function which contains many levels of nested high-order derivatives. We contribute faster and more stable AD algorithms for this challenging problem and a novel algorithm to compute exact gradients for learning. These contributions lead to significantly faster and more accurate learning algorithms, and are the first AD algorithms whose running time is polynomial in the number of levels of nesting.

Recent work has explored how to train machine learning models which do not discriminate against any subgroup of the population as determined by sensitive attributes such as gender or race. To avoid disparate treatment, sensitive attributes should not be considered. On the other hand, in order to avoid disparate impact, sensitive attributes must be examined, e.g., in order to learn a fair model, or to check if a given model is fair. We introduce methods from secure multi-party computation which allow us to avoid both. By encrypting sensitive attributes, we show how an outcome-based fair model may be learned, checked, or have its outputs verified and held to account, without users revealing their sensitive attributes.

WASHINGTON – A U.S. senator on Thursday is seeking responses from Google parent Alphabet Inc. and Twitter Inc. on whether the U.S. companies have any data-sharing agreements with Chinese vendors, following a disclosure from Facebook Inc. this week. Sen. Mark Warner, a Democrat who is vice chairman of the Intelligence Committee, said in a statement he has written letters to the companies for information on data-sharing agreements, noting that since 2012 "the relationship between the Chinese Communist Party and equipment makers like Huawei and ZTE has been an area of national security concern." Alphabet has said previously it has strategic partnerships with Chinese mobile device manufacturers, including Huawei Technologies Co. Ltd., and Xiaomi, as well as with Chinese technology platform Tencent. It wasn't clear if Twitter has a partnership with any Chinese vendors. Alphabet and Twitter did not immediately respond to questions for comment.

SAN FRANCISCO – Google will not allow its artificial intelligence software to be used in weapons or unreasonable surveillance efforts, the Alphabet Inc. unit said Thursday in standards for its business decisions in the nascent field. The new restrictions could help Google management defuse months of protest by thousands of employees against the company's work with the U.S. military to identify objects in drone video. Google will pursue other government contracts, including around cybersecurity, military recruitment and search and rescue, Chief Executive Sundar Pichai said in a blog post Thursday. "We want to be clear that while we are not developing AI for use in weapons, we will continue our work with governments and the military in many other areas," he said. Breakthroughs in the cost and performance of advanced computers have begun to carry AI from research labs into industries such as defense and health.