The new bill, called the AI Liability Directive, will add teeth to the EU's AI Act, which is set to become EU law around the same time. The AI Act would require extra checks for "high risk" uses of AI that have the most potential to harm people, including systems for policing, recruitment, or health care. The new liability bill would give people and companies the right to sue for damages after being harmed by an AI system. The goal is to hold developers, producers, and users of the technologies accountable, and require them to explain how their AI systems were built and trained. Tech companies that fail to follow the rules risk EU-wide class actions.

AI has the potential to deliver enormous business value for organizations, and its adoption has been sped up by the data-related challenges of the pandemic. Forrester estimates that almost 100% of organizations will be using AI by 2025, and the artificial intelligence software market will reach $37 billion by the same year. But there is growing concern around AI bias -- situations where AI makes decisions that are systematically unfair to particular groups of people. Researchers have found that AI bias has the potential to cause real harm. I recently had the chance to speak with Ted Kwartler, VP of Trusted AI at DataRobot, to get his thoughts on how AI bias occurs and what companies can do to make sure their models are fair.

Open & Responsible AI licenses ("OpenRAIL") are AI-specific licenses enabling open access, use and distribution of AI artifacts while requiring a responsible use of the latter. OpenRAIL licenses could be for open and responsible ML what current open software licenses are to code and Creative Commons to general content: a widespread community licensing tool. Advances in machine learning and other AI-related areas have flourished these past years partly thanks to the ubiquity of the open source culture in the Information and Communication Technologies (ICT) sector, which has permeated into ML research and development dynamics. Notwithstanding the benefits of openness as a core value for innovation in the field, (not so already) recent events related to the ethical and socio-economic concerns of development and use of machine learning models have spread a clear message: Openness is not enough. Closed systems are not the answer though, as the problem persists under the opacity of firms' private AI development processes.

Jamie Jefferson, chief creative officer at Equator, outlines the process of effective digital transformation in law firms over four concise steps. Investment is pouring into digital transformation at legal firms. Even before the pandemic, legal tech spend had reached $1.3billion according to King's College London; five times higher in 2019 than 2017. If your firm has not yet made strides towards this digital future, now is the time to act. Some companies have already automated many of their processes for the benefit of employees and clients alike.

The hard-and software that catalysed rapid developments in machine learning In late 2002 and early 2003, the release of the Radeon 9700 and GeForce FX video cards introduced a fully programmable graphics pipeline, extending and later replacing the existing fixed function pipelines. Unlike the fixed function pipeline, which allowed the user to only supply input matrices and parameters to built-in operations, the programmable pipeline introduced the execution of user-written shader programs on the GPU [Contributors, 2015]. This fundamental change allowed programmers and researchers to exploit the intrinsic parallelism of GPUs 2 years before Intel would introduce its first dual-core CPU. Within months of the availability of this new hardware and the accompanying APIs, researchers implemented linear algebra methods on GPUs and introduced programming frameworks to use GPUs for generalpurpose computations [Thompson et al., 2002, Krüger and Westermann, 2003]. This rapid development marked the dawn of general-purpose computing on graphics processing units (GPGPU). In a presentation at ICS '08, Harris presented the successes of GPGPU by highlighting a speed-up in molecular docking, N-body simulations, HD video stream transcoding, or image processing--applications in machine learning were not discussed. However, just one year later, the introduction of GPUs as general-purpose processors catalysed the deep learning explosion of the early 2010s by allowing deep learning algorithms pioneered by Alexey Ivakhnenko in 1971 to be run within practical time on widely available consumer hardware when Rajat et al. showed that GPUs outperform CPUs by an order of magnitude in large-scale deep unsupervised learning tasks [Ivakhnenko, 1971, Raina et al., 2009]. Hardware and energy requirements increase in machine learning research In 2010, Ciresan et al. [2010] introduced a multi-layer perceptron (MLP) with up to 12.11 million free parameters where forward and backward propagation were implemented on a GPU using NVIDIA's proprietary CUDA API introduced by Harris at ICS '08 two

Reasoning with legal cases, especially as conducted in common law jurisdictions such as the UK and USA, is a form of argumentation much studied in Artificial Intelligence and in computational argumentation. The formal procedure within which it is conducted and the extensive documentation which records the argument presented for each side and an assessment of these arguments make it a fruitful area for study. As described in [35], there may be several types of reasoning involved, including the use of rules, the balancing of factors, analogy and the use of policies to achieve particular purposes. All of these have been modelled in AI and Law, and this work suggests that reasoning with legal cases can been seen as going through a series of stages at which different reasoning styles are appropriate. This view will be elaborated in Section 2. One way of modelling a reasoning task [24] is to present it as a set of argumentation schemes [38]. In this paper we will use this method to articulate the reasoning required at each of the stages. Although legal reasoning is worthy of study in itself, we believe that the insights are also applicable to other, less formal, domains where it is necessary to balance reasons for and against particular options to come to a decision.

Deep learning techniques have recently been utilized for model-free age-associated gait feature extraction. However, acquiring model-free gait demands accurate pre-processing such as background subtraction, which is non-trivial in unconstrained environments. On the other hand, model-based gait can be obtained without background subtraction and is less affected by covariates. For model-based gait-based age group classification problems, present works rely solely on handcrafted features, where feature extraction is tedious and requires domain expertise. This paper proposes a deep learning approach to extract age-associated features from model-based gait for age group classification. Specifically, we first develop an unconstrained gait dataset called Multimedia University Gait Age and Gender dataset (MMU GAG). Next, the body joint coordinates are determined via pose estimation algorithms and represented as compact gait graphs via a novel part aggregation scheme. Then, a Part-AdaptIve Residual Graph Convolutional Neural Network (PairGCN) is designed for age-associated feature learning. Experiments suggest that PairGCN features are far more informative than handcrafted features, yielding up to 99% accuracy for classifying subjects as a child, adult, or senior in the MMU GAG dataset.

Businesses increasingly rely on artificial intelligence (AI) systems to make decisions that can significantly affect individual rights, human safety, and critical business operations. But how do these models derive their conclusions? What data do they use? And can we trust the results? Addressing these questions is the essence of "explainability," and getting it right is becoming essential.

Correctly labelling training data for AI models is vital to avoid serious problems, as is using sufficiently large datasets. However, manually labelling massive amounts of data is time-consuming and laborious. Using pre-labelled datasets can be problematic, as evidenced by MIT having to pull its 80 Million Tiny Images datasets. For those unaware, the popular dataset was found to contain thousands of racist and misogynistic labels that could have been used to train AI models. AI News caught up with Devang Sachdev, VP of Marketing at Snorkel AI, to find out how the company is easing the laborious process of labelling data in a safe and effective way. AI News: How is Snorkel helping to ease the laborious process of labelling data?

As a legal technology company, Legal Interact is invested in not only solving some of the legal industry's biggest challenges, but in using the best technologies and people to do so. Simply put, artificial intelligence (AI) is a branch of computer science that deals with creating intelligent computer systems, which can perform tasks that ordinarily require human intelligence. AI has already become a daily part of our lives and will eventually permeate every sphere of our professional and personal existence. At Legal Interact we have focused on an area of AI called natural language understanding (NLU) that deals with teaching computers to interpret and understand human language. Within Matter Manager, we can translate any document to English.