{"id":992148,"date":"2024-01-30T06:56:36","date_gmt":"2024-01-30T14:56:36","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-research-lab&p=992148"},"modified":"2026-09-18T16:35:04","modified_gmt":"2026-09-18T23:35:04","slug":"ai-frontiers","status":"publish","type":"msr-research-lab","link":"https:\/\/www.microsoft.com\/en-us\/research\/lab\/ai-frontiers\/","title":{"rendered":"Microsoft Research AI Frontiers"},"content":{"rendered":"
\n\t
\n\t\t
\n\t\t\t\"background\t\t<\/div>\n\t\t\n\t\t
\n\t\t\t\n\t\t\t
\n\t\t\t\t\n\t\t\t\t
\n\t\t\t\t\t\n\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\n\n

Microsoft Research
AI Frontiers<\/h2>\n\n\n\n

<\/p>\n\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t<\/div>\n\t<\/div>\n<\/section>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n

\n
\n

Who We Are<\/h3>\n\n\n\n
\n
\n

Update, Sept. 14, 2026 –<\/em> Microsoft Research AI Frontiers is now part of Microsoft AI (opens in new tab)<\/span><\/a>!<\/p>\n\n\n\n

Our lab has long focused on advancing frontier research in efficient language models and agentic systems, while reimagining existing computing through an agent-native lens. Through the years, we’ve shipped AgentInstruct<\/a>, Phi family of models (opens in new tab)<\/span><\/a>, Fara portfolio of CUA models (opens in new tab)<\/span><\/a>, the Agent Framework AutoGen (opens in new tab)<\/span><\/a>, Magentic-One<\/a>, MagenticUI<\/a>, Magentic Marketplace<\/a>. We have also led pioneering work in agent red teaming, benchmarking, and safety research. Collectively, these contributions have delivered significant value to both the academic community and the broader open-source ecosystem.<\/p>\n\n\n\n

We’ll pursue our same goal of frontier and cutting edge agentic and model research at Microsoft AI (MAI) starting now. MAI is a natural fit for our work, given MAI’s focus on humanist superintelligence and building AI systems focused on real problems and in service to people.<\/p>\n\n\n\n

We’ll share more exciting news on this transition soon!<\/p>\n<\/div>\n<\/div>\n\n\n\n

<\/div>\n\n\n\n

How We Work<\/h3>\n\n\n\n

At AI Frontiers, we combine the freedom to pursue high-impact ideas with a culture of focused, collaborative execution. We champion openness by publishing our research and contributing to open source, while driving work that advances science and delivers real-world impact. <\/p>\n\n\n\n

Our work is guided by three principles: clarity of vision<\/strong>, execution excellence<\/strong>, and deep collaboration<\/strong>. Together, these values enable us to move fast, adapt boldly, and work as One Microsoft to advance science and empower the world. <\/p>\n\n\n\n

<\/div>\n\n\n\n
\n
\n
\"background<\/figure>
\n

Our Mission<\/h3>\n\n\n\n

We believe the true transformative promise of GenAI will only happen when we start looking at the world through an agent-native lens<\/strong>.  <\/p>\n\n\n\n

This means stepping back to envision a world where agents and humans co-exist and interact seamlessly across personal and work contexts\u2014free from the constraints of existing paradigms and processes.<\/p>\n<\/div><\/div>\n<\/div>\n<\/div>\n\n\n\n

<\/p>\n\n\n\n

<\/div>\n\n\n\n

Our Research Focus<\/h3>\n\n\n\n
\n\n\n\n
\n
\n
Adaptive Agents<\/h5>\n\n\n\n

Closing the capability gap with long-running agents that learn, reason, and adapt across tasks, contexts, and modalities.<\/p>\n<\/div>\n\n\n\n

\n
Agentic Robustness<\/h5>\n\n\n\n

Closing the reliability gap to ensure consistent, trustworthy performance in real-world conditions<\/p>\n<\/div>\n\n\n\n

\n
Agentic Ecosystems<\/h5>\n\n\n\n

Reimagining workflows for an agent-native world where autonomous agents collaborate seamlessly with humans. <\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

<\/div>\n\n\n\n\n
\n\t\t
\n\t\t
\n\t\t\t