The Curious Engineer

All blogs / AI / ML

Microsoft Research

microsoft.com/en-us/research

The Microsoft Research blog provides in-depth views and perspectives from our researchers, scientists and engineers, plus announcements about noteworthy events, scholarships, and fellowships designed for academic and scientific communities.

AI / MLScienceMicrosoft blogs

’05’10’15’20’25
Posts
2,006 from 20 Jul 2001 to 30 Oct 2026
Posting
weekly (72 in the last year)
Kind
company blog
Feed
https://www.microsoft.com/en-us/research/feed/

Latest

October 2026

  1. From synthetic enterprise data evaluations to real Microsoft Copilot AI confidence
  2. Microsoft Research Focus – Edition #97
  3. Flint-interactive: A language for creating interactive data visualizations
  4. Solving a 31-year-old grand challenge in mathematics – the KLS conjecture
  5. Called to serve: Tech, research, and positive impact
  6. Proof-based general-purpose computing with Magnetar
  7. Simulating student learning for better AI tutoring
  8. Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses
  9. What AI gets wrong and what failure teaches us

September 2026

  1. Forecasting space weather risks on power grids
  2. Introducing Quine: An AI research system designed for the complexity of biology
  3. One year in: How Microsoft Research Asia – Singapore is advancing research, partnership and talent for real-world impact
  4. Offloaded inference for real-world physical AI robotics
  5. Improving synthesis prediction of small molecules at scale with RetroChimera

August 2026

  1. GigaPath-Flash and GigaTIME-Flash: Toward population-scale discovery with efficient pathology foundation models
  2. Broadening access to Skala creates a faster path to predictive DFT
  3. MindTopo reveals VLMs’ spatial reasoning abilities
  4. Introducing CARE-X: Towards Clinically Useful Radiology VLMs with Auxiliary Supervision, Reward-Aligned Learning, and Tool-Augmented Measurement
  5. Orchard: An open framework for scalable agentic AI

July 2026

  1. Echoverse: Deep, evolving environments for computer-use agents
  2. EvoLib: Turning experience into evolving knowledge
  3. Verifying Rust cryptography in SymCrypt, from standards to code
  4. Aurora 1.5: Extending open foundation models for weather and Earth-system applications
  5. Flint: A visualization language for the AI era

June 2026

  1. SkillOpt: Agent skills as trainable parameters
  2. Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity
  3. Understanding the brain with AI-driven explanations and experiments
  4. Talos: Scaling rare disease diagnosis with automated, iterative genomic reanalysis
  5. Ire identifies another LOTUSLITE specimen

May 2026

  1. Data Formulator 0.7: AI-powered data analytics for enterprise data
  2. Extending Human Intelligence Through AI
  3. MagenticLite, MagenticBrain, Fara1.5: An agentic experience optimized for small models
  4. Vega: Zero-knowledge proofs for digital identity in the age of AI
  5. Further Notes on Our Recent Research on AI Delegation and Long-Horizon Reliability
  6. mimalloc: A new, high-performance, scalable memory allocator for the modern era
  7. GridSFM: A new, small foundation model for the electric grid
  8. Advancing AI for materials with MatterSim: experimental synthesis, faster simulation, and multi-task models
  9. SocialReasoning-Bench: Measuring whether AI agents act in users’ best interests
  10. Building realistic electric transmission grid dataset at scale: a pipeline from open dataset
  11. Microsoft at NSDI 2026: Advances in large-scale networked systems

April 2026

  1. Red-teaming a network of agents: Understanding what breaks when AI agents interact at scale
  2. AutoAdapt: Automated domain adaptation for large language models
  3. Can we AI our way to a more sustainable world?
  4. New Future of Work: AI is driving rapid change, uneven benefits
  5. Ideas: Steering AI toward the work future we want
  6. ADeLe: Predicting and explaining AI performance across tasks

March 2026

  1. AsgardBench: A benchmark for visually grounded interactive planning
  2. GroundedPlanBench: Spatially grounded long-horizon task planning for robot manipulation
  3. Will machines ever be intelligent?
  4. Systematic debugging for AI agents: Introducing the AgentRx framework
  5. PlugMem: Transforming raw agent interactions into reusable knowledge
  6. Phi-4-reasoning-vision and the lessons of training a multimodal reasoning model
  7. Trailer: The Shape of Things to Come

February 2026

  1. CORPGEN advances AI agents for real work
  2. Media Authenticity Methods in Practice: Capabilities, Limitations, and Directions
  3. Project Silica’s advances in glass storage technology
  4. Rethinking imitation learning with Predictive Inverse Dynamics Models
  5. Paza: Introducing automatic speech recognition benchmarks and models for low resource languages

January 2026

  1. UniRG: Scaling medical imaging report generation with multimodal reinforcement learning
  2. Argos: Multimodal reinforcement learning with agentic verifier for AI agents
  3. OptiMind: A small language model with optimization expertise

December 2025

  1. Agent Lightning: Adding reinforcement learning to AI agents without code rewrites
  2. Promptions helps make AI prompting more precise with dynamic UI controls
  3. GigaTIME: Scaling tumor microenvironment modeling using virtual population generated by multimodal AI
  4. Ideas: Community building, machine learning, and the future of AI

November 2025

  1. Reducing Privacy leaks in AI: Two approaches to contextual integrity
  2. Fara-7B: An Efficient Agentic Model for Computer Use
  3. MMCTAgent: Enabling multimodal reasoning over large video and image collections
  4. BlueCodeAgent: A blue teaming agent enabled by automated red teaming for CodeGen AI
  5. When industry knowledge meets PIKE-RAG: The innovation behind Signify’s customer service boost
  6. Magentic Marketplace: an open-source simulation environment for studying agentic markets
  7. RedCodeAgent: Automatic red-teaming agent against diverse code agents

October 2025

  1. Tell me when: Building agents that can wait, monitor, and act
  2. Ideas: More AI-resilient biosecurity with the Paraphrase Project
  3. When AI Meets Biology: Promise, Risk, and Responsibility

September 2025

  1. Using AI to assist in rare disease diagnosis
  2. Tool-space interference in the MCP era: Designing for agent compatibility at scale
  3. RenderFormer: How neural networks are reshaping 3D rendering
  4. Breaking the networking wall in AI infrastructure

August 2025

  1. Crescent library brings privacy to digital identity systems
  2. Applicability vs. job displacement: further notes on our recent research on AI and occupations
  3. Coauthor roundtable: Reflecting on healthcare economics, biomedical research, and medical education
  4. MindJourney enables AI to explore simulated 3D worlds to improve spatial interpretation
  5. Dion: the distributed orthonormal update revolution is here
  6. Reimagining healthcare delivery and public health with AI
  7. Self-adaptive reasoning for science
  8. Project Ire autonomously identifies malware at scale
  9. VeriTrail: Detecting hallucination and tracing provenance in multi-step AI workflows

July 2025

  1. Navigating medical education in the era of generative AI
  2. Xinxing Xu bridges AI research and real-world impact at Microsoft Research Asia – Singapore
  3. Technical approach for classifying human-AI interactions at scale
  4. AI Testing and Evaluation: Reflections
  5. CollabLLM: Teaching LLMs to collaborate with users
  6. AI Testing and Evaluation: Learnings from cybersecurity
  7. How AI will accelerate biomedical research and discovery
  8. AI Testing and Evaluation: Learnings from pharmaceuticals and medical devices

June 2025

  1. AI Testing and Evaluation: Learnings from genome editing
  2. PadChest-GR: A bilingual grounded radiology reporting benchmark for chest X-rays
  3. AI Testing and Evaluation: Learnings from Science and Industry
  4. Learning from other domains to advance AI evaluation and testing
  5. Breaking bonds, breaking ground: Advancing the accuracy of computational chemistry with deep learning
  6. New methods boost reasoning in small and large language models
  7. How AI is reshaping the future of healthcare and medical research
  8. Rewriting SymCrypt in Rust to modernize Microsoft’s cryptographic library
  9. BenchmarkQED: Automated benchmarking of RAG systems

May 2025

  1. What AI’s impact on individuals means for the health workforce and industry
  2. FrodoKEM: A conservative quantum-safe cryptographic algorithm
  3. Abstracts: Zero-shot models in single-cell biology with Alex Lu
  4. Abstracts: Aurora with Megan Stanley and Wessel Bruinsma
  5. Collaborators: Healthcare Innovation to Impact
  6. Magentic-UI, an experimental human-centered web agent
  7. Coauthor roundtable: Reflecting on real world of doctors, developers, patients, and policymakers
  8. Predicting and explaining AI model performance: A new approach to evaluation
  9. Abstracts: Heat Transfer and Deep Learning with Hongxia Hao and Bing Lv
  10. Research Focus: Week of May 7, 2025
  11. Microsoft Fusion Summit explores how AI can accelerate fusion research
  12. Abstracts: Societal AI with Xing Xie
  13. Societal AI: Building human-centered AI systems
  14. Laws, norms, and ethics for AI in health

April 2025

  1. Research Focus: Week of April 21, 2025
  2. The Future of AI in Knowledge Work: Tools for Thought at CHI 2025
  3. Empowering patients and healthcare consumers in the age of generative AI
  4. Engagement, user expertise, and satisfaction: Key insights from the Semantic Telemetry Project
  5. Debug-gym: an environment for AI coding tools to learn how to debug code like programmers
  6. Research Focus: Week of April 7, 2025
  7. Real-world healthcare AI development and deployment—at scale
  8. VidTok introduces compact, efficient tokenization to enhance AI video processing

March 2025

  1. Ideas: Accelerating Foundation Models Research: AI for all
  2. Research Focus: Week of March 24, 2025
  3. The reality of generative AI in the clinic
  4. Metasurface: Unlocking the future of wireless sensing and communication
  5. Claimify: Extracting high-quality claims from language model outputs
  6. Introducing KBLaM: Bringing plug-and-play external knowledge to LLMs
  7. Semantic Telemetry: Understanding how users interact with AI systems
  8. The AI Revolution in Medicine, Revisited: An Introduction
  9. Advancing biomedical discovery: Overcoming data challenges in precision medicine

February 2025

  1. Magma: A foundation model for multimodal AI agents across digital and physical worlds
  2. Exploring the structural changes driving protein function with BioEmu-1
  3. Introducing Muse: Our first generative AI model designed for gameplay ideation
  4. Ideas: Quantum computing redefined with Chetan Nayak
  5. Microsoft Research and Physics Wallah team up to enhance AI-based tutoring
  6. ExACT: Improving AI agents’ decision-making via test-time compute scaling
  7. Ideas: Building AI for population-scale systems with Akshay Nambi
  8. Advances to low-bit quantization enable LLMs on edge devices

January 2025

  1. Research Focus: Week of January 27, 2025
  2. Ideas: Bug hunting with Shan Lu
  3. Research Focus: Week of January 13, 2025
  4. Ideas: AI for materials discovery with Tian Xie and Ziheng Lu
  5. MatterGen: A new paradigm of materials design with generative AI
  6. AutoGen v0.4: Reimagining the foundation of agentic AI for scale, extensibility, and robustness

Every year

2026 612025 892024 1372023 1202022 482021 782020 1062019 1452018 1422017 822016 822015 1172014 1622013 1762012 1582011 1402010 712009 232008 172007 132006 132005 72004 182001 1