AI BriefingMicrosoft Research news
The latest AI and developer news from Microsoft Research, with short summaries.
Visit website →NEWMicrosoft Research Focus #98: KLS Conjecture Solved and Coding Agent Costs Cut by 25.4%
The edition highlights a proof of the KLS conjecture and a new method, APEC, that reduces coding agent processing costs by up to 25.4%.
Microsoft Researchers Discover Lazarus: Backdoor Technique Reprogramming Dormant Neurons in AI Agents
The attack requires no retraining and passes standard safety evaluations, with triggers as simple as a date change.
PDD Kit uses Verus to formally verify agent-written Rust web apps
A research prototype from Microsoft Research and GitHub Next that shifts review from line-by-line code to plain-language intent, using formal proofs to ensure agent-written Rust code matches specifications.
Microsoft Research Manager Mike Walker: Security Rules Are Social Constructs, and AI Has Upended the Cost of Network Attacks
Walker argues that the 2013-2016 Cyber Grand Challenge relied on expert systems rather than machine learning, and that current AI capabilities have disrupted the traditional balance of offensive scarcity.
Microsoft establishes first standardized evaluation baseline for Copilot using synthetic enterprise data
The platform uses publicly available data sources like O*NET and SEC filings to generate realistic test scenarios without using customer data.
Microsoft Research Focus Edition 97: ForkPilot cuts agent processing by 59%, BioEmu2 enters public preview
BioEmu2, a foundation model for biomolecular structure, is now available in public preview, while new frameworks improve coding agents and weather forecasts.
Microsoft Research introduces Flint-interactive for AI-generated interactive data visualizations
The system uses a high-level language and compiler to reduce the cost and complexity of creating interactive visualizations with AI agents.
Microsoft Research Proves 31-Year-Old KLS Conjecture Using Human-Directed AI
Researchers proved the 1995 Kannan-Lovász-Simonovits conjecture, establishing a dimension-free geometric principle for log-concave distributions.
Microsoft Research releases Magnetar, a hash-based zkVM proving Rust and Go programs 10-39x faster
Magnetar proves RISC-V execution at 7.2 to 8.7 million instructions per second on a laptop using only standard hash functions.
Microsoft Research introduces StudentSim for evaluating AI tutors with personalized student simulators
StudentSim outperforms baselines like GPT-5.4 and Maia2 in chess, writing, and math simulations, achieving 90.5% tutor accuracy in reinforcement learning tests.
Microsoft Research Asia Releases Agent Lightning v1.0: A 3,500-Line Framework for Harnessed Agentic RL
Microsoft Research Asia released Agent Lightning v1.0, a 3,500-line framework that raised Qwen3.5-9B's Pass@1 on SWE-bench Verified by 14.6 percentage points using only 6,000 training samples.
Microsoft Research Researchers Discuss 'Surprising Failures' in Multi-Turn AI Conversations
Microsoft Research scientists Jennifer Neville and Chad Atalla discuss how AI models degrade in multi-turn interactions despite high single-turn benchmark scores, highlighting the need for new evaluation methods and user strategies.
Microsoft Research ML pipeline forecasts space-weather risk for 66,935 US substations
The system provides 30 to 60 minutes of advance warning and detected nearly 80% of major space-weather events during evaluation.
Microsoft Research Introduces Quine, a Multimodal AI System for Biological Discovery
Quine combines a multimodal world model with an interactive harness to prioritize therapeutic compounds, validated in collaboration with the Broad Institute.
Microsoft Research Asia – Singapore Marks One Year with Expanded AI Partnerships and Talent Programs
The lab has launched nine new projects with NUS and NTU and engaged over 300 students through its summer school since opening in July 2025.
MS Enhances Battery Life and Performance by Offloading Physical AI Robotics Inference
MS unveiled Physical AI Toolchain features that offload robotics inference to the edge or cloud, extending battery life by up to 160% and improving performance.
Nature-published retrosynthesis AI 'RetroChimera' released… Demonstrates superior performance over existing models in expert evaluation
Microsoft has open-sourced the code and weights of RetroChimera, a retrosynthesis AI published in Nature, under the MIT license.
Chris White: From DARPA and Dark Web Search to Microsoft Catalyst Lab: Solving Social Problems with Technology
Chris White leads AI security and infrastructure innovation at Microsoft Catalyst Lab after developing dark web search tools at DARPA.
GigaPath-Flash and GigaTIME-Flash Released: Efficient Pathology Foundation Models Accelerate Large-Scale Research
Flash models maximizing the efficiency of GigaPath and GigaTIME have been released, enhancing accessibility for large-scale pathology research.
Expanded Skala Accessibility: A Faster Path to Predictive DFT
Microsoft Research released Skala 1.1 with improved accuracy and expanded integration with major simulation codes.