Designing safe digital systems for the humanitarian sector

The Illusion of Inclusion How LLMs Misrepresent African Languages and Cultural Contexts

Data Formulator v0.5: Vibe with data, in control

Knowledge-Coin Fair Exchange

Distant conversational speech recognition: Challenges and Opportunities

Ultra Ethernet for next-generation AI and HPC workloads

BRAIN SIGNALS TO ACTION: Monitoring and Explaining User Cognitive Load with Foundation Models

IronDict: Transparent Dictionaries from Polynomial Commitments

Lattice-Based Accumulator and Application to Anonymous Credential Revocation

FOA Tokenizer: Learning Discrete Representations of Spatial Audio with Multichannel VQ-GAN

Efficient Secure Aggregation for Federated Learning

What does it mean to paraphrase a protein?

Evaluating the Cultural Relevance of AI Models and Products: Insights from the YUX Team

Evaluating Privacy Policies under Modern Privacy Laws At Scale: An LLM-Based Automated Approach

Ideas: More AI-resilient biosecurity with the Paraphrase Project

Detecting Compromise of Passkey Storage on the Cloud

Encrypted Access Logging for Online Accounts: Device Attributions without Device Tracking

Microsoft Research India - The evolution

Microsoft Research India - The lab culture

Understanding How Users Prepare for and React to Smartphone Theft

When LLMs Go Online: The Emerging Threat of Web-Enabled LLMs

A Formal Analysis of Apple’s iMessage PQ3 Protocol

Email Spoofing with SMTP Smuggling: How the Shared Email Infrastructures Magnify this Vulnerability

A Framework for Abusability Analysis: The Case of Passkeys in Interpersonal Threat Models

‘Hey mum, I dropped my phone down the toilet’: Investigating Hi Mum and Dad SMS Scams in the UK

Dehumanizing machines: Making sense of AI systems that seem human

Scalable emulation of protein equilibrium ensembles with BioEmu

Disrupting the AI infrastructure with MicroLEDs

Dion: The distributed orthonormal update revolution is here

Pushing boundaries of complex reasoning in small language models

Microsoft Research Forum | Season 2, Episode 1

zk-promises: Anonymous Moderation, Reputation, & Blocking from Anonymous Credentials with Callbacks

More is Less: Extra Features in Contactless Payments Break Security

Sub-Population Identification of Multi-morbidity in Sub-Saharan African Populations

Echoes in GenAI generations

Six Years of Rowhammer: Breakthroughs and Future Directions

MindJourney: Test-Time Scaling with World Models for Spatial Reasoning

Medical Bayesian Kiosk (2010)

Reimagining healthcare delivery and public health with AI

VeriTrail: Detect hallucination and trace provenance in AI workflows

Computational models for brain science

VoluMe: Authentic 3D Video Calls from Live Gaussian Splat Prediction

How I became a StoryTeller (and how YOU can too)

Make some noise: Teaching the language of audio to an LLM using sound tokens

Building Better Language Models Through Global Understanding

Navigating medical education in the era of generative AI

DAViD: Data-efficient and Accurate Vision Models from Synthetic Data

AI Testing and Evaluation: Reflections

Intern talk: Distilling Self-Supervised-Learning-Based Speech Quality Assessment into Compact Models

AI Testing and Evaluation: Learnings from cybersecurity

Scalable emulation of protein equilibrium ensembles with BioEmu

How AI will accelerate biomedical research and discovery

Introducing Microsoft AI Economy Institute

AI Testing and Evaluation: Learnings from pharmaceuticals and medical devices

Against Softmaxing Culture: Understanding Relational Practices in Expert and Ordinary Forms of Work

AI Testing and Evaluation: Learnings from genome editing

AI Testing and Evaluation: Learnings from Science and Industry

Precio: Private Aggregate Measurement via Oblivious Shuffling

Sandi: A System for Accountability

DFT for drug and material discovery