Apple on Thursday said it's introducing new child safety features in iOS, iPadOS, watchOS, and macOS as part of its efforts to limit the spread of Child Sexual Abuse Material (CSAM) in the U.S. To that effect, the iPhone maker said it intends to begin client-side scanning of images shared via every Apple device for known child abuse content as they are being uploaded into iCloud Photos, in addition to leveraging on-device machine learning to vet all iMessage images sent or received by minor accounts (aged under 13) to warn parents of sexually explicit photos shared over the messaging platform. Furthermore, Apple also plans to update Siri and Search to stage an intervention when users try to perform searches for CSAM-related topics, alerting that the "interest in this topic is harmful and problematic." "Messages uses on-device machine learning to analyze image attachments and determine if a photo is sexually explicit," Apple noted. "The feature is designed so that Apple does not get access to the messages." The feature, called Communication Safety, is said to be an opt-in setting that must be enabled by parents through the Family Sharing feature. Detection of known CSAM images involves carrying out on-device matching using a database of known CSAM image hashes provided by the National Center for Missing and Exploited Children (NCMEC) and other child safety organizations before the photos are uploaded to the cloud.

The large and still increasing popularity of deep learning clashes with a major limit of neural network architectures, that consists in their lack of capability in providing human-understandable motivations of their decisions. In situations in which the machine is expected to support the decision of human experts, providing a comprehensible explanation is a feature of crucial importance. The language used to communicate the explanations must be formal enough to be implementable in a machine and friendly enough to be understandable by a wide audience. In this paper, we propose a general approach to Explainable Artificial Intelligence in the case of neural architectures, showing how a mindful design of the networks leads to a family of interpretable deep learning models called Logic Explained Networks (LENs). LENs only require their inputs to be human-understandable predicates, and they provide explanations in terms of simple First-Order Logic (FOL) formulas involving such predicates. LENs are general enough to cover a large number of scenarios. Amongst them, we consider the case in which LENs are directly used as special classifiers with the capability of being explainable, or when they act as additional networks with the role of creating the conditions for making a black-box classifier explainable by FOL formulas. Despite supervised learning problems are mostly emphasized, we also show that LENs can learn and provide explanations in unsupervised learning settings. Experimental results on several datasets and tasks show that LENs may yield better classifications than established white-box models, such as decision trees and Bayesian rule lists, while providing more compact and meaningful explanations.

We review practical challenges in building and deploying ethical AI at the scale of contemporary industrial and societal uses. Apart from the purely technical concerns that are the usual focus of academic research, the operational challenges of inconsistent regulatory pressures, conflicting business goals, data quality issues, development processes, systems integration practices, and the scale of deployment all conspire to create new ethical risks. Such ethical concerns arising from these practical considerations are not adequately addressed by existing research results. We argue that a holistic consideration of ethics in the development and deployment of AI systems is necessary for building ethical AI in practice, and exhort researchers to consider the full operational contexts of AI systems when assessing ethical risks.

Regulators have signalled an interest in adopting explainable AI(XAI) techniques to handle the diverse needs for model governance, operational servicing, and compliance in the financial services industry. In this short overview, we review the recent technical literature in XAI and argue that based on our current understanding of the field, the use of XAI techniques in practice necessitate a highly contextualized approach considering the specific needs of stakeholders for particular business applications.

We introduce the needs for explainable AI that arise from Standard No. 239 from the Basel Committee on Banking Standards (BCBS 239), which outlines 11 principles for effective risk data aggregation and risk reporting for financial institutions. Of these, explainableAI is necessary for compliance in two key aspects: data quality, and appropriate reporting for multiple stakeholders. We describe the implementation challenges for one specific regulatory requirement:that of having a complete data taxonomy that is appropriate for firmwide use. The constantly evolving nature of financial ontologies necessitate a continuous updating process to ensure ongoing compliance.

Defeasible rules are used in providing computable representations of legal documents and, more recently, have been suggested as a basis for explainable AI. Such applications draw attention to the scalability of implementations. The defeasible logic $DL(\partial_{||})$ was introduced as a more scalable alternative to $DL(\partial)$, which is better known. In this paper we consider the use of (implementations of) $DL(\partial_{||})$ as a computational aid to computing conclusions in $DL(\partial)$ and other defeasible logics, rather than as an alternative to $DL(\partial)$. We identify conditions under which $DL(\partial_{||})$ can be substituted for $DL(\partial)$ with no change to the conclusions drawn, and conditions under which $DL(\partial_{||})$ can be used to draw some valid conclusions, leaving the remainder to be drawn by $DL(\partial)$.

Product liability cases have dogged Amazon in recent years as it emerged as the nation's largest online retailer, in part by turning its store into an online bazaar fueled by millions of third-party vendors. By prioritizing vast selection, the company has allowed merchants to sell on the site with scant vetting. And while the company defends its screening processes, which uses machine learning technology, among other processes, to identify risky sellers, it has faced suits across the country over defective items that hurt customers and destroyed their property.

Delivering justice is an important responsibility which requires a person to have flawless knowledge of the law and an approach completely free of all human biases. Apart from the pressure of avoiding errors that could be devastating for those involved in legal disputes, the judiciary also has to deal with the cases piling up, in order to ensure swift functioning of courts. Smart tech has started replacing traditional methods in all sectors including crime fighting and transportation, so that human error can be avoided while achieving higher efficiency. Following Dubai Police's use of facial recognition and brainwaves for investigation, courts in the futuristic city are also embracing AI, which will replace judges and resolve real estate as well as rental issues. Designed to speed up the judicial process in the Emirate, the smart court mechanism will cut out biases and tackle loopholes, to increase accuracy in litigation.

All the sessions from Transform 2021 are available on-demand now. Joonko has raised $10 million to create a prequalified pool of diverse job candidates for corporate recruiters. The Tel Aviv, Israel-based human resources tech startup is built on a remarkably simple idea. Lots of companies are recruiting diverse candidates in the wake of Black Lives Matter, #MeToo, and other social movements. But they often can hire only one person amid a pool of diverse candidates.

The California Department of Fair Employment and Housing sounds like a bureaucratic borefest, but it's actually pretty important. It files lawsuits against companies and landlords accused of discrimination. Today we talk about California's lawsuit against Activision Blizzard. The Santa Monica company made $8 billion last year on the strength of classic video game titles like "Call of Duty" and "World of Warcraft." But the state argues the company let fester a "pervasive frat boy workplace culture" that led to sexual harassment against women.