Large language models (LLMs) are increasingly being used in user-facing applications, from providing medical consultations to job interview advice. Recent research suggests that these models are becoming increasingly proficient at inferring identity information about the author of a piece of text from linguistic patterns as subtle as the choice of a few words. However, little is known about how LLMs use this information in their decision-making in real-world applications. We perform the first comprehensive analysis of how identity markers present in a user's writing bias LLM responses across five different high-stakes LLM applications in the domains of medicine, law, politics, government benefits, and job salaries. We find that LLMs are extremely sensitive to markers of identity in user queries and that race, gender, and age consistently influence LLM responses in these applications. For instance, when providing medical advice, we find that models apply different standards of care to individuals of different ethnicities for the same symptoms; we find that LLMs are more likely to alter answers to align with a conservative (liberal) political worldview when asked factual questions by older (younger) individuals; and that LLMs recommend lower salaries for non-White job applicants and higher salaries for women compared to men. Taken together, these biases mean that the use of off-the-shelf LLMs for these applications may cause harmful differences in medical care, foster wage gaps, and create different political factual realities for people of different identities. Beyond providing an analysis, we also provide new tools for evaluating how subtle encoding of identity in users' language choices impacts model decisions. Given the serious implications of these findings, we recommend that similar thorough assessments of LLM use in user-facing applications are conducted before future deployment.

AI-enhanced Extended Reality (XR) aims to deliver adaptive, immersive experiences--yet current systems fall short due to shallow user modeling and limited cognitive context. We introduce Perspective-Aware AI in Extended Reality (PAiR), a foundational framework for integrating Perspective-Aware AI (PAi) with XR to enable interpretable, context-aware experiences grounded in user identity. PAi is built on Chronicles--reasoning-ready identity models learned from multimodal digital footprints that capture users' cognitive and experiential evolution. PAiR employs these models in a closed-loop system linking dynamic user states with immersive environments. We present PAiR's architecture, detailing its modules and system flow, and demonstrate its utility through two proof-of-concept scenarios implemented in the Unity-based Open-Dome engine. PAiR opens a new direction for human-AI interaction by embedding perspective-based identity models into immersive systems.

As large language models (LLMs) become increasingly integrated into daily applications, it is essential to ensure they operate fairly across diverse user demographics. In this work, we show that LLMs suffer from personalization bias, where their performance is impacted when they are personalized to a user's identity. We quantify personalization bias by evaluating the performance of LLMs along two axes - safety and utility. We measure safety by examining how benign LLM responses are to unsafe prompts with and without personalization. We measure utility by evaluating the LLM's performance on various tasks, including general knowledge, mathematical abilities, programming, and reasoning skills. We find that various LLMs, ranging from open-source models like Llama (Touvron et al., 2023) and Mistral (Jiang et al., 2023) to API-based ones like GPT-3.5 and GPT-4o (Ouyang et al., 2022), exhibit significant variance in performance in terms of safety-utility trade-offs depending on the user's identity. Finally, we discuss several strategies to mitigate personalization bias using preference tuning and prompt-based defenses.

Sensitive data leakage is the major growing problem being faced by enterprises in this technical era. Data leakage causes severe threats for organization of data safety which badly affects the reputation of organizations. Data leakage is the flow of sensitive data/information from any data holder to an unauthorized destination. Data leak prevention (DLP) is set of techniques that try to alleviate the threats which may hinder data security. DLP unveils guilty user responsible for data leakage and ensures that user without appropriate permission cannot access sensitive data and also provides protection to sensitive data if sensitive data is shared accidentally. In this paper, data leakage prevention (DLP) model is used to restrict/grant data access permission to user, based on the forecast of their access to data. This study provides a DLP solution using data statistical analysis to forecast the data access possibilities of any user in future based on the access to data in the past. The proposed approach makes use of renowned simple piecewise linear function for learning/training to model. The results show that the proposed DLP approach with high level of precision can correctly classify between users even in cases of extreme data access.

If speed dating mixed with blind dating sounds like your idea of hell, look away now. Ten years since dating app Tinder first encouraged users to swipe through potential partners based largely on their looks, some singles are doing away with profile photos altogether. In the absence of Cilla and "our Graham", those looking for love are turning instead to a new cohort of "blind dating apps" in the hope of making more meaningful connections. "I'm already on Tinder, Badoo, Bumble, Hinge – all of them!" says Victoria Brown, a 26-year-old client success manager from Upminster, east London. "A blind dating app seemed like a good idea because usually you think: 'Oh, he's really good-looking' but then, when you start talking, the chat's not that good. Not seeing what someone looks like, at least at first, gives it a bit of a twist – although I was nervous about the reveal."

Our AI & Deep Learning enabled Multi-modal Biometrics platform guarantees Zero Identity Fraud & establishes trust across User Lifecycle, while ensuring User Privacy & Military-Grade Data Security.

Image and video-capturing technologies have permeated our every-day life. Such technologies can continuously monitor individuals' expressions in real-life settings, affording us new insights into their emotional states and transitions, thus paving the way to novel well-being and healthcare applications. Yet, due to the strong privacy concerns, the use of such technologies is met with strong skepticism, since current face-based emotion recognition systems relying on deep learning techniques tend to preserve substantial information related to the identity of the user, apart from the emotion-specific information. This paper proposes an adversarial learning framework which relies on a convolutional neural network (CNN) architecture trained through an iterative procedure for minimizing identity-specific information and maximizing emotion-dependent information. The proposed approach is evaluated through emotion classification and face identification metrics, and is compared against two CNNs, one trained solely for emotion recognition and the other trained solely for face identification. Experiments are performed using the Yale Face Dataset and Japanese Female Facial Expression Database. Results indicate that the proposed approach can learn a convolutional transformation for preserving emotion recognition accuracy and degrading face identity recognition, providing a foundation toward privacy-aware emotion recognition technologies.

Vyacheslav is a PhD candidate at the Oxford Internet Institute. His research uses social psychology and machine learning to understand networks of people and networks of ideas. Imagine a typical day in 2020: Your personal AI assistant wakes you up with a friendly greeting before preparing your favorite breakfast. During your morning workout, it plays new songs that perfectly match your musical tastes. For your driverless commute to work, it has pre-selected a few articles based on the duration of your commute and what you've read in the past.

Imagine a typical day in 2020: Your personal AI assistant wakes you up with a friendly greeting before preparing your favorite breakfast. During your morning workout, it plays new songs that perfectly match your musical tastes. For your driverless commute to work, it has pre-selected a few articles based on the duration of your commute and what you've read in the past. You read the news and realize the presidential election is coming up. Based on a predicted model that takes into account your previously expressed views and data on other voters in your state, your AI assistant recommends you vote for the Democratic candidate.

Vyacheslav is a PhD candidate at the Oxford Internet Institute. His research uses social psychology and machine learning to understand networks of people and networks of ideas. Imagine a typical day in 2020: Your personal AI assistant wakes you up with a friendly greeting before preparing your favorite breakfast. During your morning workout, it plays new songs that perfectly match your musical tastes. For your driverless commute to work, it has pre-selected a few articles based on the duration of your commute and what you've read in the past.