Large language models (LLMs) are susceptible to a variety of risks, from non-faithful output to biased and toxic generations. Due to several limiting factors surrounding LLMs (training cost, API access, data availability, etc.), it may not always be feasible to impose direct safety constraints on a deployed model. Therefore, an efficient and reliable alternative is required. To this end, we present our ongoing efforts to create and deploy a library of detectors: compact and easy-to-build classification models that provide labels for various harms. In addition to the detectors themselves, we discuss a wide range of uses for these detector models - from acting as guardrails to enabling effective AI governance. We also deep dive into inherent challenges in their development and discuss future work aimed at making the detectors more reliable and broadening their scope.

Explaining neural model predictions to users requires creativity. Especially in enterprise applications, where there are costs associated with users' time, and their trust in the model predictions is critical for adoption. For link prediction in master data management, we have built a number of explainability solutions drawing from research in interpretability, fact verification, path ranking, neuro-symbolic reasoning and self-explaining AI. In this demo, we present explanations for link prediction in a creative way, to allow users to choose explanations they are more comfortable with.

Although the rise of Large Language Models (LLMs) in enterprise settings brings new opportunities and capabilities, it also brings challenges, such as the risk of generating inappropriate, biased, or misleading content that violates regulations and can have legal concerns. To alleviate this, we present "LLMGuard", a tool that monitors user interactions with an LLM application and flags content against specific behaviours or conversation topics. To do this robustly, LLMGuard employs an ensemble of detectors.

The prevalence of half-truths, which are statements containing some truth but that are ultimately deceptive, has risen with the increasing use of the internet. To help combat this problem, we have created a comprehensive pipeline consisting of a half-truth detection model and a claim editing model. Our approach utilizes the T5 model for controlled claim editing; "controlled" here means precise adjustments to select parts of a claim. Our methodology achieves an average BLEU score of 0.88 (on a scale of 0-1) and a disinfo-debunk score of 85% on edited claims. Significantly, our T5-based approach outperforms other Language Models such as GPT2, RoBERTa, PEGASUS, and Tailor, with average improvements of 82%, 57%, 42%, and 23% in disinfo-debunk scores, respectively. By extending the LIAR PLUS dataset, we achieve an F1 score of 82% for the half-truth detection model, setting a new benchmark in the field. While previous attempts have been made at half-truth detection, our approach is, to the best of our knowledge, the first to attempt to debunk half-truths.

We propose a method to control the attributes of Language Models (LMs) for the text generation task using Causal Average Treatment Effect (ATE) scores and counterfactual augmentation. We explore this method, in the context of LM detoxification, and propose the Causally Fair Language (CFL) architecture for detoxifying pre-trained LMs in a plug-and-play manner. Our architecture is based on a Structural Causal Model (SCM) that is mathematically transparent and computationally efficient as compared with many existing detoxification techniques. We also propose several new metrics that aim to better understand the behaviour of LMs in the context of toxic text generation. Further, we achieve state of the art performance for toxic degeneration, which are computed using \RTP (RTP) benchmark. Our experiments show that CFL achieves such a detoxification without much impact on the model perplexity. We also show that CFL mitigates the unintended bias problem through experiments on the BOLD dataset.

Data exploration and quality analysis is an important yet tedious process in the AI pipeline. Current practices of data cleaning and data readiness assessment for machine learning tasks are mostly conducted in an arbitrary manner which limits their reuse and results in loss of productivity. We introduce the concept of a Data Readiness Report as an accompanying documentation to a dataset that allows data consumers to get detailed insights into the quality of input data. Data characteristics and challenges on various quality dimensions are identified and documented keeping in mind the principles of transparency and explainability. The Data Readiness Report also serves as a record of all data assessment operations including applied transformations. This provides a detailed lineage for the purpose of data governance and management. In effect, the report captures and documents the actions taken by various personas in a data readiness and assessment workflow. Overtime this becomes a repository of best practices and can potentially drive a recommendation system for building automated data readiness workflows on the lines of AutoML [8]. We anticipate that together with the Datasheets [9], Dataset Nutrition Label [11], FactSheets [1] and Model Cards [15], the Data Readiness Report makes significant progress towards Data and AI lifecycle documentation.

We used Python's Flask framework for building the service and exposed a REST API that generates a bias report based on the following input parameters from a user: the dataset name, the protected attributes, the privileged and unprivileged groups, the chosen fairness metrics, and the chosen mitigation algorithm, if any. With these inputs, the back-end then runs a series of steps to 1) split the dataset into training, development, and validation sets; 2) train a logistic regression classifier on the training set; 3) run the bias-checking metrics on the classifier against the test dataset; 4) if a mitigation algorithm is chosen, run the mitigation algorithm with the appropriate pipeline (pre-processing, in-processing, or post-processing). The end result is then cached so that if the exact same inputs are provided, the result can be directly retrieved from cache and no additional computation is needed. The reason to truly use the toolkit code in serving the Web application rather than having a pre-computed lookup table of results is twofold: we want to make the app a real representation of the underlying capabilities (in fact, creating the Web app helped us debug a few items in the code), and we also avoid any issues of synchronizing updates to the metrics, explainers, and algorithms with the results shown: synchronization is automatic. Currently, the service is limited to three built-in datasets, but it can be expanded to support the user's own data upload. The service is also limited to building logistic regression classifiers, but again this can be expanded. Such expansions can be more easily implemented if this fairness service is integrated into a full AI suite that provides various classifier options and data storage solutions.

Machine Learning community is recently exploring the implications of bias and fairness with respect to the AI applications. The definition of fairness for such applications varies based on their domain of application. The policies governing the use of such machine learning system in a given context are defined by the constitutional laws of nations and regulatory policies enforced by the organizations that are involved in the usage. Fairness related laws and policies are often spread across the large documents like constitution, agreements, and organizational regulations. These legal documents have long complex sentences in order to achieve rigorousness and robustness. Automatic extraction of fairness policies, or in general, any specific kind of policies from large legal corpus can be very useful for the study of bias and fairness in the context of AI applications. We attempted to automatically extract fairness policies from publicly available law documents using two approaches based on semantic relatedness. The experiments reveal how classical Wordnet-based similarity and vector-based similarity differ in addressing this task. We have shown that similarity based on word vectors beats the classical approach with a large margin, whereas other vector representations of senses and sentences fail to even match the classical baseline. Further, we have presented thorough error analysis and reasoning to explain the results with appropriate examples from the dataset for deeper insights.

The accuracy and reliability of machine learning algorithms are an important concern for suppliers of artificial intelligence (AI) services, but considerations beyond accuracy, such as safety, security, and provenance, are also critical elements to engender consumers' trust in a service. In this paper, we propose a supplier's declaration of conformity (SDoC) for AI services to help increase trust in AI services. An SDoC is a transparent, standardized, but often not legally required, document used in many industries and sectors to describe the lineage of a product along with the safety and performance testing it has undergone. We envision an SDoC for AI services to contain purpose, performance, safety, security, and provenance information to be completed and voluntarily released by AI service providers for examination by consumers. Importantly, it conveys product-level rather than component-level functional testing. We suggest a set of declaration items tailored to AI and provide examples for two fictitious AI services.

The presence of gender stereotypes in many aspects of society is a well-known phenomenon. In this paper, we focus on studying and quantifying such stereotypes and bias in the Man Bookers Prize winning fiction. We consider 275 books shortlisted for Man Bookers Prize between 1969 and 2017. The gender bias is analyzed by semantic modeling of book descriptions on Goodreads. This reveals the pervasiveness of gender bias and stereotype in the books on different features like occupation, introductions and actions associated to the characters in the book.