Microsoft Research

Three white line icons on a blue-to-green gradient background: a computer monitor with a globe symbol on the left, a cursor arrow with click lines in the center, and a computer mouse outline on the right.

Fara-7B is our first agentic small language model for computer use. This experimental model includes robust safety measures to aid responsible deployment. Despite its size, Fara-7B holds its own against larger, more resource-intensive agentic systems: msft.it/6015tpZHF

Coming December 9 at 8:00 AM PT, our last Microsoft Research Forum episode of the year. 

Register now: www.microsoft.com/en-us/resear...

Project Gecko is bridging the AI equity gap by creating cost-effective, culturally relevant AI systems that deliver expertise and insight to the global majority. This Microsoft Research-led initiative is empowering communities with AI-driven solutions that drive better outcomes: msft.it/6011trtP7

Project Gecko brings AI to underserved populations reflecting local language and culture, Workload Intelligence enhances cloud platform efficiency, and more. msft.it/6013tRExc

If you’re eager to test, validate, and influence what’s next in enterprise AI, this session is your launchpad.

Ashley Llorens, Marco Casalaina, and Shannon Monroe will provide a sneak peek into the future of AI with Foundry Labs, your front door to experimental models, multi-agent orchestration prototypes, Agent Factory blueprints, and edge innovations.

Ready for #MSIgnite ~ Join us for AI Frontier in Foundry Labs: Experiment Today, Lead Tomorrow on Wed, Nov 19 from 4:00 PM to 4:45 PM PST online and in San Francisco (Moscone West, Level 3, Room 3011). msft.it/6015tO23S

Magentic Marketplace is an open-source simulation environment for studying how AI agents interact and transact in digital markets. Its modular design lets researchers experiment with different market setups, agent behaviors, and system dynamics at scale. msft.it/6014tMvva

How to stop AI from killing your critical thinking Chatbots might help you get work done faster — but at what cost? When we outsource our reasoning to artificial intelligence, we reduce ourselves to "middle managers for our own thoughts," says AI and ...

His talk explored how AI transforms knowledge work, creativity, and human thought – and how, if we design AI to make us think, and not think for us, we can greatly benefit as a society. Check out Advait’s presentation to learn more: msft.it/6018tOP38

How to stop AI from killing your critical thinking Chatbots might help you get work done faster — but at what cost? When we outsource our reasoning to artificial intelligence, we reduce ourselves to "middle managers for our own thoughts," says AI and ...

Microsoft senior researcher, Advait Sarkar, recently presented at TED AI Vienna, where he spoke about AI’s role in elevating human thought.

A field of green plants with two people walking in between the rows.

AI tools often perform poorly in non-Western languages and lack cultural context that’s crucial for many populations. Project Gecko uses small language models to deliver vital content that extends AI’s benefits to serve the global majority: msft.it/6018t3Rrn

Three white line icons on a blue-to-green gradient background: the first icon on the left shows a network with a central circle connected to smaller circles; the middle icon shows a microchip with a shield symbol in the center; the right icon depicts a flowchart with connected shapes.

BlueCodeAgent is an end-to-end blue-teaming framework built to boost code security using automated red-teaming processes, data, and safety rules to guide LLMs’ defensive decisions. Dynamic testing reduces false positives in vulnerability detection: msft.it/6014tMvnr

Headshot of Esha Choukse near text: Esha Choukse awarded the ACM SIGMICRO Early Career Award. A blue to pink gradient background sits behind text and the headshot.

Congratulations to Esha Choukse for receiving the ACM SIGMICRO Early Career Award! This recognition honors her for foundational contributions to hardware memory compression and sustainable and efficient datacenter systems. She is a Principal Researcher in Azure Research – Systems. msft.it/6014tzZsC

Check out Trista Chen, Flavio Griggio, Ahmed Awadallah, Jina Suh, and Cecily Morrison's conversations with Forward Future, a daily AI newsletter, to learn more about the people building the systems that are becoming part of our daily lives.

Behind the breakthroughs: A day in the life at Microsoft Research | Microsoft Signal Blog | Microsoft Five Microsoft researchers shared their stories with Forward Future, a daily AI newsletter, about the big problems they’re tackling and the trends shaping their work in the middle of an ever-changing ...

Microsoft researchers are driving trends and tackling major challenges in the midst of the ever-changing AI era. msft.it/6015tJvdx

Data Formulator v0.5: Vibe with data, in control - Microsoft Research Data Formulator is an AI-powered visualization tool that helps analysts iteratively explore and visualize data from any source—clean or messy, small or large—without coding. In agent mode, users provi...

To get started with Agentic Mode, watch a full tutorial video here: msft.it/6013tJyvb 2/2

Take your data exploration to the next level with Agentic Mode, now in the newest version of Data Formulator, available on Azure AI Foundry Labs. Update charts, get recommendations, and create reports grounded in your exploration. msft.it/6010tJyvk 1/2

Three white line icons on a gradient background transitioning from blue to purple to peach. From left to right: a network diagram with a central circle connected to smaller circles, an icon of two people inside a rectangular frame, and a magnifying glass.

A collaboration between Signify and Microsoft Research shows how PIKE-RAG improves enterprise knowledge systems, delivering a 12% increase in accuracy and faster, more reliable answers: msft.it/6012tydu8

AI agents are transforming digital marketplaces, mediating discovery and transactions between consumers and businesses. The new Magentic Marketplace provides an open-source, extensible simulation environment for studying different agentic market designs: msft.it/6014tyz9q

Icons of a chat bubble, connected document, and shield with checkmark on a blue-green gradient background.

Code agents help streamline software development workflows, but may also introduce critical security risks. Learn how RedCodeAgent automates and improves “red-teaming” attack simulations to help uncover real-world threats that other methods overlook: msft.it/6013shouB

In this issue of Research Focus: ECHO boosts LM agents’ learning efficiency; Robusta enhances heuristic algorithms with LLMs; LEGOMem improves multi-agent workflows; and PulseParse secures data parsing: msft.it/6016tGwXK

The paper explores the challenges of collaboration between software engineers and domain experts in fields like AI, health, and science. Congratulations to the team!

Microsoft researchers received the Best Paper Award at ESEM 2025 for "When Domains Collide: An Activity Theory Exploration of Cross-Disciplinary Collaboration.” Learn more: msft.it/6012tEVKm

The StarTrack Scholars Program at Microsoft Research Asia offers young faculty a 3-month research opportunity to collaborate globally, gain access to world-class resources, and innovate for the future. Learn more and apply now: msft.it/6012t8Sme

Research is serious work, powered by real people. Get to know Director of Microsoft Research AI for Science Chris Bishop and learn what excites him beyond the lab.

Workflow icons showing tasks, thinking, and time, linked to a person shaped symbol on a gradient background.

SentinelStep enables AI agents to handle monitoring tasks that run for hours or days, like watching for emails or tracking prices. It works by managing when agents should check and their context, avoiding wasted resources and missed updates. msft.it/6018sC85Y

Paraphrasing a protein is like paraphrasing a sentence—the words change, but the meaning holds. Microsoft scientists applied this concept to toxic proteins, revealing a blind spot in DNA screening and leading a cross-sector team to develop a fix. msft.it/6010s4tW8

Illustrated headshots of Eric Horvitz, Bruce Wittmann, Tessa Alexanian, and James Diggans. Ideas: A Microsoft Research Podcast runs along the bottom of the graphic.

Identifying and mitigating risk to support AI breakthroughs needs thoughtful, innovative approaches. In “Ideas,” Eric Horvitz and collaborators share the nuances of building a red-teaming protocol for testing and securing DNA biosecurity screening tools. msft.it/6015sF1vt

We’re a proud sponsor of #SOSP2025! From advancing performance debugging, to multilingual AI systems, our teams are pushing the boundaries of systems research. Explore the full list of accepted papers: msft.it/6014sNzS8

How is agentic AI changing research in the modern lab? Join our October 28 webinar to explore real use cases in life sciences and see how AI copilots boost R&D productivity. msft.it/6019sFEZb