This quarter marked demonstrable progress toward sector-specific approaches to the regulation of artificial intelligence and machine learning ("AI"). As the EU continues to inch toward finalizing its draft Artificial Intelligence Act--the landmark, cross-sector regulatory framework for AI/ML technologies--the White House published a "Blueprint for an AI Bill of Rights," a non-binding set of principles memorializing the Biden administration's approach to algorithmic regulation. The AI Bill of Rights joins a number of recent U.S. legislative proposals, both at the federal and state levels,[1] and the Federal Trade Commission's ("FTC") Advanced Notice of Proposed Rulemaking to solicit input on questions related to potentially harmful data privacy and security practices, including automated decision-making systems. Our 3Q22 Artificial Intelligence and Automated Systems Legal Update focuses on these regulatory efforts and also examines other policy developments within the U.S. and Europe. The past several years have seen a number of new algorithmic governance initiatives take shape at the federal level, building on the December 2020 Trustworthy AI Executive Order that outlined nine distinct principles to ensure agencies "design, develop, acquire and use AI in a manner that fosters public trust and confidence while protecting privacy."[2]

Smart sensors, devices and systems deployed in smart cities have brought improved physical protections to their citizens. Enhanced crime prevention, and fire and life safety protection are achieved through these technologies that perform motion detection, threat and actors profiling, and real-time alerts. However, an important requirement in these increasingly prevalent deployments is the preservation of privacy and enforcement of protection of personal identifiable information. Thus, strong encryption and anonymization techniques should be applied to the collected data. In this IEEE Big Data Cup 2022 challenge, different masking, encoding and homomorphic encryption techniques were applied to the images to protect the privacy of their contents. Participants are required to develop detection solutions to perform privacy preserving matching of these images. In this paper, we describe our solution which is based on state-of-the-art deep convolutional neural networks and various data augmentation techniques. Our solution achieved 1st place at the IEEE Big Data Cup 2022: Privacy Preserving Matching of Encrypted Images Challenge.

In federated learning (FL), classifiers (e.g., deep networks) are trained on datasets from multiple centers without exchanging data across them, and thus improves sample efficiency. In the classical setting of FL, the same labeling criterion is usually employed across all centers being involved in training. This constraint greatly limits the applicability of FL. For example, standards used for disease diagnosis are more likely to be different across clinical centers, which mismatches the classical FL setting. In this paper, we consider an important yet under-explored setting of FL, namely FL with mixed-type labels where different labeling criteria can be employed by various centers, leading to inter-center label space differences and challenging existing FL methods designed for the classical setting. To effectively and efficiently train models with mixed-type labels, we propose a theory-guided and model-agnostic approach that can make use of the underlying correspondence between those label spaces and can be easily combined with various FL methods such as FedAvg. We present convergence analysis based on over-parameterized ReLU networks. We show that the proposed method can achieve linear convergence in label projection, and demonstrate the impact of the parameters of our new setting on the convergence rate. The proposed method is evaluated and the theoretical findings are validated on benchmark and medical datasets.

Machine learning (ML) models have been deployed for high-stakes applications, e.g., healthcare and criminal justice. Prior work has shown that ML models are vulnerable to attribute inference attacks where an adversary, with some background knowledge, trains an ML attack model to infer sensitive attributes by exploiting distinguishable model predictions. However, some prior attribute inference attacks have strong assumptions about adversary's background knowledge (e.g., marginal distribution of sensitive attribute) and pose no more privacy risk than statistical inference. Moreover, none of the prior attacks account for class imbalance of sensitive attribute in datasets coming from real-world applications (e.g., Race and Sex). In this paper, we propose an practical and effective attribute inference attack that accounts for this imbalance using an adaptive threshold over the attack model's predictions. We exhaustively evaluate our proposed attack on multiple datasets and show that the adaptive threshold over the model's predictions drastically improves the attack accuracy over prior work. Finally, current literature lacks an effective defence against attribute inference attacks. We investigate the impact of fairness constraints (i.e., designed to mitigate unfairness in model predictions) during model training on our attribute inference attack. We show that constraint based fairness algorithms which enforces equalized odds acts as an effective defense against attribute inference attacks without impacting the model utility. Hence, the objective of algorithmic fairness and sensitive attribute privacy are aligned.

Extensive efforts have been made to understand and improve the fairness of machine learning models based on observational metrics, especially in high-stakes domains such as medical insurance, education, and hiring decisions. However, there is a lack of certified fairness considering the end-to-end performance of an ML model. In this paper, we first formulate the certified fairness of an ML model trained on a given data distribution as an optimization problem based on the model performance loss bound on a fairness constrained distribution, which is within bounded distributional distance with the training distribution. We then propose a general fairness certification framework and instantiate it for both sensitive shifting and general shifting scenarios. In particular, we propose to solve the optimization problem by decomposing the original data distribution into analytical subpopulations and proving the convexity of the subproblems to solve them. We evaluate our certified fairness on six real-world datasets and show that our certification is tight in the sensitive shifting scenario and provides non-trivial certification under general shifting. Our framework is flexible to integrate additional non-skewness constraints and we show that it provides even tighter certification under different real-world scenarios. We also compare our certified fairness bound with adapted existing distributional robustness bounds on Gaussian data and demonstrate that our method is significantly tighter.

While it may take many by surprise, that's the fresh call to action among analysts paying close attention to how companies are – or aren't – factoring artificial intelligence (AI) and machine learning (ML) into their data management plans and playbooks. After years of reading sensational stories about the limitless potential of intelligent machines, stakeholders and C-suite, executives in particular appear to be confused about the best course of action to take. Commercial missteps and the total failure of some products have resulted. Experts say it doesn't have to be this way. "AI and ML has become crucial and necessary for nearly all businesses in every sector," says Elliott Young, CTO, Dell Technologies UK. "In the same way that businesses have had to transform digitally and become digital-first, companies are going to need AI and ML to remain competitive. Those on the path towards this are already reaping the benefits of being able to make decisions driven by predictive analytics."

Data, algorithms and compute (the main elements of AI) have advanced rapidly over the past two decades and are being deployed in people's lives in disruptive ways -- for good and ill -- without much regulation or, until recently, community deliberation. The fallout is considerable, and many concerns have been raised over harms of AI algorithms in society to individuals or groups. A recent example is a white paper by the American Civil Liberties Union (ACLU) entitled "AI in healthcare may worsen medical racism"4. The paper discusses several studies that inadvertently used biased data and machine learning models to make harmful medical decisions. In recent years, the machine learning community has called for participatory approaches whereby those whose life is impacted by algorithms have a major role in their development5.

Reveal-Brainspace announced that Intellectual Data, an eDiscovery service provider in Korea, will be integrating Reveal's AI-powered eDiscovery, review & investigations platform – Reveal 11 – onto its suite of enterprise cloud services for legal and corporate entities throughout the region. Specifically, Intellectual Data will leverage Reveal's end-to-end, SaaS-based platform to offer eDiscovery hosting, business process optimization and consulting services to its clients – all underpinned by advanced AI and machine learning technology. "As the first partner of Reveal in Korea, we look forward to collaborating on evolving the service to mitigate the risk factors blocking global growth of our clients." "Korea is a lynchpin in Reveal's strategic growth initiative in the APAC region, which makes the partnership with one of Korea's most respected eDiscovery service providers even more significant," said Wendell Jisa, CEO of Reveal. "We're thrilled to work with the talented team at Intellectual Data to provide to expand the reach of our Reveal 11 platform to the growing network of law firms and corporations in Korea looking to solve their most complex challenges with leading AI and review technology."

Managing expenses isn't a task that many people particularly love to work on, yet it's a vital one and can have enormous ramifications for a business if not handled properly. If you'd like to find a way to spend less time sorting out this area of your organizational finances and reduce problems simultaneously, it's worth turning to technology. Thanks to artificial intelligence (AI), expense management is becoming easier and more effective than ever. In many organizations, a lack of space or too much paperwork to store leads to a lot of stress and time spent attempting to locate files. If this is a pain point in your venture, AI can help at least limit expense-based paperwork, among other things.

If you know anything about Ghost Robotics, it's likely one of two things: 1) They make robot dogs. A majority of the Philadelphia firm's press coverage has revolved around these facts, along with some coverage of its systems being used to patrol the U.S. border. It's shameful how both parties fight tooth nail to defend their ability to pump endless public money into militarization. From tanks in police depts to corrupt military contracts, funding this violence is bipartisan non-controversial, yet healthcare housing isn't. Ghost has thus far not demonstrated any manner of ethical qualms when it comes to its work with military and law enforcement -- but it's the company's product design that could ultimately get it in hot water.