Artificial intelligence, as a capability enhancer, offers significant improvements to our tactical warfighting advantage. AI provides methods for fusing and analyzing data to enhance our knowledge of the tactical environment; it provides methods for generating and assessing decision options from multidimensional, complex situations; and it provides predictive analytics to identify and examine the effects of tactical courses of action. Machine learning can improve these processes in an evolutionary manner. Advanced computing techniques can handle highly heterogeneous and vast datasets and can synchronize knowledge across distributed warfare assets. This article presents concepts for applying AI to various aspects of tactical battle management and discusses their potential improvements to future warfare.

Traditional cyber security techniques have led to an asymmetric disadvantage for defenders. The defender must detect all possible threats at all times from all attackers and defend all systems against all possible exploitation. In contrast, an attacker needs only to find a single path to the defender’s critical information. In this article, we discuss how this asymmetry can be rebalanced using cyber deception to change the attacker’s perception of the network environment, and lead attackers to false beliefs about which systems contain critical information or are critical to a defender’s computing infrastructure. We introduce game theory concepts and models to represent and reason over the use of cyber deception by the defender and the effect it has on attacker perception. Finally, we discuss techniques for combining artificial intelligence algorithms with game theory models to estimate hidden states of the attacker using feedback through payoffs to learn how best to defend the system using cyber deception. It is our opinion that adaptive cyber deception is a necessary component of future information systems and networks. The techniques we present can simultaneously decrease the risks and impacts suffered by defenders and dramatically increase the costs and risks of detection for attackers. Such techniques are likely to play a pivotal role in defending national and international security concerns.

The AI Bookie column documents highlights from AI Bets, an online forum for the creation of adjudicatable predictions, in the form of bets, about the future of AI. While it is easy to make broad, generalized, or off-the-cuff predictions about the future, it is more difficult to develop predictions that are carefully thought out, concrete, and measurable. This forum was created to help researchers craft predictions whose accuracy can be clearly and unambiguously judged when the bets come due. The bets will be documented both online and regularly in this column. We encourage bets that are rigorously and scientifically argued. We discourage bets that are too general to be evaluated or too specific to an individual or institution. The goal is not to continue to feed the media frenzy and outsized pundit predictions about AI, but rather to curate and promote bets whose outcomes will provide useful feedback to the scientific community. For detailed guidelines and to place bets, visit sciencebets.org.

The purpose of this article is to draw attention to an aspect of intelligence that has not yet received significant attention from the AI community, but that plays a crucial role in a technology’s effectiveness in the world, namely teaming intelligence. We propose that Al will reach its full potential only if, as part of its intelligence, it also has enough teaming intelligence to work well with people. Although seemingly counterintuitive, the more intelligent the technological system, the greater the need for collaborative skills. This paper will argue why teaming intelligence is important to AI, provide a general structure for AI researchers to use in developing intelligent systems that team well, assess the current state of the art and, in doing so, suggest a path forward for future AI systems. This is not a call to develop a new capability, but rather, an approach to what AI capabilities should be built, and how, so as to imbue intelligent systems with teaming competence.

XAI, explainable AI, and the DARPA program to provide that. The European Commission has legislated a demand (Order GDPR 2016/2679) specifying that deployed machine learning systems must explain their decisions. The commission has done this even though no one knows how to provide what they are requiring. What would follow if we and machines are in roughly the same position with respect to the transparency of our ethical decision-making? I want to reintroduce the notion of orthosis into ethical explanation: medically, an orthosis is an externally applied device designed and fitted to the body to aid rehabilitation, and usually contrasted with a prosthesis, which replaces a missing part, like a foot or leg. Here, it will mean an explanatory software agent associated with a human or machine.

Defense and security organizations depend upon science and technology to meet operational needs, predict and counter threats, and meet increasingly complex demands of modern warfare. Artificial intelligence and robotics could provide solutions to a wide range of military gaps and deficiencies. At the same time, the unique and rapidly evolving nature of AI and robotics challenges existing polices, regulations, and values, and introduces complex ethical issues that might impede their development, evaluation, and use by the Canadian Armed Forces (CAF). Early consideration of potential ethical issues raised by military use of emerging AI and robotics technologies in development is critical to their effective implementation. This article presents an ethics assessment framework for emerging AI and robotics technologies. It is designed to help technology developers, policymakers, decision makers, and other stakeholders identify and broadly consider potential ethical issues that might arise with the military use and integration of emerging AI and robotics technologies of interest. We also provide a contextual environment for our framework, as well as an example of how our framework can be applied to a specific technology. Finally, we briefly identify and address several pervasive issues that arose during our research.

Autonomous Vehicles (AV) are expected to bring considerable benefits to society, such as traffic optimization and accidents reduction. They rely heavily on advances in many Artificial Intelligence (AI) approaches and techniques. However, while some researchers in this field believe AI is the core element to enhance safety, others believe AI imposes new challenges to assure the safety of these new AI-based systems and applications. In this non-convergent context, this paper presents a systematic literature review to paint a clear picture of the state of the art of the literature in AI on AV safety. Based on an initial sample of 4870 retrieved papers, 59 studies were selected as the result of the selection criteria detailed in the paper. The shortlisted studies were then mapped into six categories to answer the proposed research questions. An AV system model was proposed and applied to orient the discussions about the SLR findings. As a main result, we have reinforced our preliminary observation about the necessity of considering a serious safety agenda for the future studies on AI-based AV systems.

Adversarial examples are carefully constructed modifications to an input that completely change the output of a classifier but are imperceptible to humans. Despite these successful attacks for continuous data (such as image and audio samples), generating adversarial examples for discrete structures such as text has proven significantly more challenging. In this paper we formulate the attacks with discrete input on a set function as an optimization task. We prove that this set function is submodular for some popular neural network text classifiers under simplifying assumption. This finding guarantees a $1-1/e$ approximation factor for attacks that use the greedy algorithm. Meanwhile, we show how to use the gradient of the attacked classifier to guide the greedy search. Empirical studies with our proposed optimization scheme show significantly improved attack ability and efficiency, on three different text classification tasks over various baselines. We also use a joint sentence and word paraphrasing technique to maintain the original semantics and syntax of the text. This is validated by a human subject evaluation in subjective metrics on the quality and semantic coherence of our generated adversarial text.

Adversarial examples are important for understanding the behavior of neural models, and can improve their robustness through adversarial training. Recent work in natural language processing generated adversarial examples by assuming white-box access to the attacked model, and optimizing the input directly against it (Ebrahimi et al., 2018). In this work, we show that the knowledge implicit in the optimization procedure can be distilled into another more efficient neural network. We train a model to emulate the behavior of a white-box attack and show that it generalizes well across examples. Moreover, it reduces adversarial example generation time by 19x-39x. We also show that our approach transfers to a black-box setting, by attacking The Google Perspective API and exposing its vulnerability. Our attack flips the API-predicted label in 42\% of the generated examples, while humans maintain high-accuracy in predicting the gold label.

This paper considers the problem of Byzantine fault tolerance in distributed linear regression in a multi-agent system. However, the proposed algorithms are given for a more general class of distributed optimization problems, of which distributed linear regression is a special case. The system comprises of a server and multiple agents, where each agent is holding a certain number of data points and responses that satisfy a linear relationship (could be noisy). The objective of the server is to determine this relationship, given that some of the agents in the system (up to a known number) are Byzantine faulty (aka. actively adversarial). We show that the server can achieve this objective, in a deterministic manner, by robustifying the original distributed gradient descent method using norm based filters, namely 'norm filtering' and 'norm-cap filtering', incurring an additional log-linear computation cost in each iteration. The proposed algorithms improve upon the existing methods on three levels: i) no assumptions are required on the probability distribution of data points, ii) system can be partially asynchronous, and iii) the computational overhead (in order to handle Byzantine faulty agents) is log-linear in number of agents and linear in dimension of data points. The proposed algorithms differ from each other in the assumptions made for their correctness, and the gradient filter they use.