Insider Brief

• Microsoft is using agentic AI throughout its quantum computing program to accelerate the development of its Majorana 2 chip, automate research workflows and support its goal of building a commercially useful quantum computer by 2029.
• The company says AI agents help scientists analyze nearly two decades of research data, identify patterns across disciplines, manage complex engineering dependencies and uncover manufacturing issues that might otherwise go unnoticed.
• Microsoft reports that agentic AI has significantly reduced experimental cycle times by automating measurements, optimizing materials research and enabling researchers to test and refine quantum device designs more efficiently.
• Image: At Microsoft’s Quantum Lab in Lyngby, Denmark, scientists, software engineers and fabrication experts are using agentic AI across many of their processes to speed the development of more reliable topological qubits. (Microsoft)

As quantum AI — the idea that quantum computers could one day enhance artificial intelligence — steadily climbs the list of anticipated quantum computing use cases, Microsoft says AI is already playing a growing role in advancing its quantum computing program.

The company said it is increasingly relying on agentic artificial intelligence to accelerate development of its quantum computing program, using AI systems to help design materials, automate experiments, analyze decades of research data and identify problems that human researchers might miss.

According to a recent Microsoft blog post, the company’s quantum team has integrated agentic AI across much of its work on Majorana 2, Microsoft’s latest quantum chip, as the company pursues a goal of building a commercially useful quantum computer by 2029.

 

The effort reflects a broader shift underway across scientific research, where AI is moving beyond writing assistance and data analysis into a more active role. Rather than simply answering questions, agentic AI systems can carry out multi-step tasks, propose hypotheses, coordinate workflows and continuously learn from results while remaining under human supervision.

Microsoft executives say those capabilities are helping overcome some of the most difficult engineering challenges facing quantum computing.

“Agentic AI has permeated almost everything we do—it’s just become kind of a very natural part of our workflow,” Nayak said in the post. “The agents can really accelerate things as much or as little as you want. It can be as little as pulling information together and summarizing it, or it can go further down the road of synthesizing it more or generating an interesting hypothesis. I think that’s extremely powerful right now.”

The company’s quantum ambitions center on topological qubits, a type of quantum bit designed to be more resistant to errors than conventional approaches. Quantum computers store information in qubits, which can exist in multiple states simultaneously, potentially allowing them to solve certain problems far faster than traditional computers.

The problem is that qubits are notoriously fragile. Small disturbances from heat, radiation or almost any type of environmental noise can cause calculations to fail.

According to Microsoft, Majorana 2 represents a substantial improvement in reliability. The company said the chip’s qubits can maintain their quantum state roughly 1,000 times longer than those used in the first-generation Majorana system introduced last year.

Microsoft reported mean qubit lifetimes of about 20 seconds, with some devices lasting as long as one minute. For comparison, many competing quantum systems measure qubit lifetimes in microseconds.

The company attributes part of that progress to advances in materials science and engineering, but executives say AI has become an increasingly important tool in accelerating those improvements.

AI Becomes Part of the Scientific Process

Alongside the Majorana 2 announcement, Microsoft announced the general availability of Microsoft Discovery, a research platform designed to support scientific and engineering work through specialized AI agents.

According to the company, the platform combines autonomous AI agents with research workflows, enterprise security controls and reasoning systems intended to help scientists manage complex projects.

Researchers can deploy teams of AI agents that analyze large collections of information, generate hypotheses, optimize experiments and validate theories while working under human oversight.

Microsoft said scientists inside its quantum program have already been using many of those capabilities.

The company’s approach highlights how AI is increasingly being positioned as a collaborator rather than a replacement for researchers.

Microsoft executives repeatedly emphasized what they describe as a “scientist in the loop” model, where human researchers retain responsibility for decisions while AI helps process information and suggest directions.

Managing Complex Quantum Development

Quantum computing projects involve a vast web of interconnected disciplines.

Researchers must simultaneously manage software, chip architecture, materials science, fabrication processes, measurement systems and device physics. A change in one area can affect performance elsewhere.

According to Microsoft, agentic AI helps researchers track those relationships and identify patterns across systems that would be difficult for individuals to detect on their own.

The company’s quantum effort has also accumulated nearly two decades of research data stored across numerous formats and repositories.

Before the introduction of modern AI systems, much of that information remained isolated in separate databases and teams.

“As you run AI agents on this data, they’re able to essentially resynthesize and make correlations that we as humans cannot see because no single individual has that much vision across that much data,” Zulfi Alam, corporate vice president for quantum at Microsoft, said in the post.

The challenge becomes even greater because Microsoft’s quantum researchers are spread across multiple countries and specialties, including physics, materials science, mechanical engineering and manufacturing.

To address that problem, Microsoft developed AI systems capable of synthesizing information across disciplines and presenting recommendations to researchers.

“The AI is able to synthesize knowledge from all these different disciplines,” Alam added.

Accelerating Experiments and Materials Discovery

Microsoft says some of the most significant gains have come from applying AI to experimental work.

Creating and operating topological qubits requires researchers to adjust hundreds of variables before measurements can begin. Historically, those processes could take weeks.

According to Alam, Microsoft previously attempted to automate parts of that workflow using earlier machine learning systems but encountered limitations.

However, using capabilities now available through Microsoft Discovery, the team developed specialized AI agents capable of automating measurements and exploring large numbers of parameter combinations simultaneously.

Microsoft said the approach reduced experimental cycle times by orders of magnitude.

The AI systems can continuously adjust voltages, collect data and build detailed maps of operating conditions while searching for optimal configurations.

“Using agentic AI to automate the measurements was a game changer,” Alam said. “It goes through some math and starts saying, ‘Hey, where do I find the lowest point where everything sort of works?’ And it can do all these voltage adjustments in parallel, which a human cannot do. The way our minds work, we are more linear.”

The technology is also being applied to materials development.

Majorana 2 uses a different superconducting material than the first-generation Majorana chip. While Majorana 1 relied on aluminum, Microsoft switched to lead in the new design.

According to Nayak, the materials change produced significant improvements in device performance, though it required extensive research to overcome engineering tradeoffs.

Future materials work is expected to make even greater use of AI-driven simulations.

Alam said researchers can use agentic AI to identify likely material configurations through simulation before conducting physical experiments, potentially reducing lengthy trial-and-error cycles.

“Finding the exact recipe, the right amount to put to get the desired energy structure, requires a lot of experimentation in the old world order,” Alam said. “In the new world order, through simulations, you can see where the highly probable target is. And then with that knowledge, you ideally only have to experiment once.”

Beyond Quantum Computing

Microsoft reports that the lessons learned from its quantum program can extend to scientific research more broadly.

The company said organizations in industries including life sciences, chemicals, materials development, manufacturing and energy are already using Microsoft Discovery for research and development projects.

Aseem Datar, corporate vice president for product innovation at Microsoft Discovery, said customers have begun exploring applications ranging from semiconductor manufacturing materials to industrial research challenges.

The broader goal is to create a research environment where scientists can examine larger datasets, test more ideas and coordinate more disciplines than would otherwise be possible.

For Microsoft, the quantum program serves as one of the earliest demonstrations of that strategy.

The company still faces significant challenges before reaching its goal of a scalable quantum computer. But according to Microsoft executives, agentic AI is increasingly becoming part of the development process itself, helping researchers move faster through some of the most difficult scientific and engineering problems.

Alam added that this fundamentally new type of Frontier R&D lets a scientist, “be the anchor point and look at many, many different disciplines all at the same time with a very high fidelity and be able to draw correlations from that. It is the essence of what every single high-performance, cutting-edge team wants to do.”

AI-Powered Quantum: Microsoft Turns to Agentic AI to Speed Quantum Computing Push