Microsoft Advocates for Open Standards to Enable Cross-Company AI Agent Collaboration

Microsoft wants AI 'agents' to work together and remember things

REDMOND, Washington, May 18 (Reuters) - Microsoft envisions a future where any company's artificial intelligence agents can work together with agents from other firms and have better memories of their interactions, its chief technologist said on Sunday ahead of the company's annual software developer conference. Microsoft is holding its Build conference in Seattle on May 19, where analysts expect the company to unveil its latest tools for developers building AI systems. Speaking at Microsoft's headquarters in Redmond, Washington, ahead of the conference, Chief Technology Officer Kevin Scott told reporters and analysts the company is focused on helping spur the adoption of standards across the technology industry that will let agents from different makers collaborate. Agents are AI systems that can accomplish specific tasks, such as fixing a software bug, on their own. Scott said that Microsoft is backing a technology called Model Context Protocol (MCP), an open-source protocol introduced by Google-backed Anthropic. Scott said MCP has the potential to create an "agentic web" similar to the way hypertext protocols that helped spread the internet in the 1990s. "It means that your imagination gets to drive what the agentic web becomes, not just a handful of companies that happen to see some of these problems first," Scott said. Scott also said that Microsoft is trying to help AI agents have better memories of things that users have asked them to do, noting that, so far, "most of what we're building feels very transactional." But making an AI agent's memory better costs a lot of money because it requires more computing power. Microsoft is focusing on a new approach called structured retrieval augmentation, where an agent extracts short bits of each turn in a conversation with a user, creating a roadmap to what was discussed. "This is a core part of how you train a biological brain - you don't brute force everything in your head every time you need to solve a particular problem," Scott said. Reporting by Stephen Nellis; Editing by Paul Simao Our Standards: The Thomson Reuters Trust Principles., opens new tab Suggested Topics:Artificial Intelligence

What AI can learn from the browser wars: Microsoft CTO calls for open standards to fuel 'agentic web'

REDMOND, Wash. -- Microsoft is pitching a future built on the "agentic web," where AI agents don't just answer questions but move across platforms, access tools, and carry out sophisticated tasks autonomously. But the company's current leaders are also reflecting on the past -- taking lessons from the early days of the web, in which a prior version of Microsoft initially took a different approach. Speaking with reporters and analysts Sunday night, ahead of the company's Build developer conference this week, Microsoft CTO Kevin Scott urged the industry to prioritize openness and simplicity. He warned that a closed, vertically integrated approach could stifle innovation. Scott said AI agents will only reach their full potential if they can "talk to everything in the world." If early browser makers had been able to tightly control the web through vertically integrated protocols, focusing on their own priorities, it would have been a "less interesting version of the web than the one that we have right now," Scott said. During the internet's rapid expansion in the 1990s, Microsoft drew antitrust scrutiny over its integration of Internet Explorer into Windows, with critics accusing the company of trying to control the web through proprietary protocols and anti-competitive tactics. Fast-forward to 2025, and Google's Chrome dominates global browser usage, giving the search giant significant influence over web standards. The U.S. Department of Justice wants to force Google to divest Chrome as part of a broader antitrust case against the company. Microsoft, for its part, now bases its Edge browser on Chromium, Google's open-source engine -- shifting from dominating the browser market to joining a shared ecosystem, following a steep drop in its market share. Google, which is holding its own I/O developer conference this week, recently introduced the Agent-to-Agent protocol, or A2A, an open standard designed to let AI agents communicate and collaborate directly across systems. Scott briefly mentioned A2A during his remarks Sunday night, but focused more on the Model Context Protocol, or MCP. Developed and open-sourced by Anthropic -- the AI startup backed by Amazon and Google -- MCP defines how AI models access and interact with external tools and data sources. Microsoft's key AI partner, OpenAI, is also reportedly working to integrate MCP into ChatGPT. "MCP in particular, we're very, very excited about," Scott said, citing its simplicity and flexibility. He compared MCP to foundational web protocols like HTTP -- simple by design but flexible enough to support future innovation. He said its structure allows developers to build on it over time, adding new capabilities without replacing the entire system. AI reasoning models -- systems capable of planning, decision-making, and solving multi-step problems -- are progressing fast, Scott said, but they still need reliable ways to connect with the digital world. MCP is "filling such an unbelievably big need in the ecosystem," he said. "Even though what Anthropic put out early is a really, really super simple implementation, it's really kind of breathtaking." Still, he said, solving these problems will require open collaboration. "The important thing about them isn't that Microsoft's opinion gets expressed, or some big tech company's opinion gets expressed, or that we've won some kind of technical argument," he said. What's important, he said, is to "get to the thing that's really ubiquitous as quickly as humanly possible." Wrapping up his talk, Scott said the current era reminds him of the early days of the web, when simple tools enabled broad experimentation without requiring permission from gatekeepers. "That's the moment you all are in right now," he said. "All of these building blocks are available." Expect to hear more on these themes Monday morning as the Build conference opens with a keynote from Scott and Microsoft CEO Satya Nadella at the Seattle Convention Center. Stay tuned for coverage on GeekWire.

Microsoft wants everyone to use an open-source technology to create an 'agentic web' where AI agents interact with other AI agents

In the spirit of a technology developed by AI company Anthropic, Microsoft sees the future of AI where there are lots of different systems, created by lots of different companies, all working together, in peace and harmony. Or to put it in the same words that Microsoft used, create an "agentic web". That's according to a report by Reuters, which relayed the views of Microsoft's chief technology officer, Kevin Scott, ahead of the software company's annual Build conference. What Scott hopes to achieve is to have Microsoft's AI agents happily work with those from other companies, via a standard platform called the Model Context Protocol (MCP). This is an open-source standard, created by Anthropic -- an AI business that's a mere four years old. The idea behind it is that it makes it much easier for AI systems to access and share for training purposes, and when it comes to the specific area of AI agents, it should help them perform far better at their tasks. Fundamentally, AI agents are a type of artificial intelligence system that just do one very specific task, such as searching through code for a certain bug and then fixing it. They run autonomously, analysing data and then make a decision based on rules set out during the AI's training. Agentic AI, to use the proper name for it all, has a wide range of potential applications, such as cybersecurity and customer support, but it's only as good as the data it has been trained on. Enter stage left, MCP, which essentially lets AI agents work hand-in-hand (or should that be tensor-in-tensor?) with other agents to improve the accuracy of their outputs. According to Reuters, Scott remarked that "MCP has the potential to create an 'agentic web' similar to the way hypertext protocols helped spread the internet in the 1990s." It's not just about training data, though, as the accuracy of agentic AI depends heavily on something called reinforcement learning. Similar to how 'rewards' and 'punishments' affect the behaviour of animals, reinforcement learning helps AI agents focus on optimising their outputs based on achieving the biggest rewards. Having AI agents share what works and what doesn't would certainly be useful in reinforcement learning, but it doesn't raise the question as to what happens if agents are simply left to their own devices. Does one simply assume that the network of agents will never accidentally pick a negative strategy over a positive one? What mechanisms would need to be created to prevent an 'agentic web' from spirally into a negative feedback loop? Better brains than mine will surely have raised the same questions by now and, hopefully, developed systems to prevent all of this from happening. In the same way that certain stocks and shares are automatically sold and bought by computers, with next to no human interaction at all, we could be nearing the point where many aspects of our lives are decided for us by an enormous network of interconnected AI agents. For example, customer support services for banks, emergency services, and other vital systems could well be entirely agentic AI within a decade or so. I'm not knowledgeable enough about AI to sensibly judge if this is a really good thing or a really bad one, but my gut feelings suggest that the reality of the situation will end up being somewhere between the two extremes. Let's just hope is much closer to the former than the latter, yes?

Microsoft wants AI 'agents' to work together and remember things

Speaking at Microsoft's headquarters in Redmond, Washington, ahead of the conference, Chief Technology Officer Kevin Scott told reporters and analysts the company is focused on helping spur the adoption of standards across the technology industry that will let agents from different makers collaborate. Agents are AI systems that can accomplish specific tasks, such as fixing a software bug, on their own.Microsoft envisions a future where any company's artificial intelligence agents can work together with agents from other firms and have better memories of their interactions, its chief technologist said on Sunday ahead of the company's annual software developer conference. Microsoft is holding its Build conference in Seattle on May 19, where analysts expect the company to unveil its latest tools for developers building AI systems. Speaking at Microsoft's headquarters in Redmond, Washington, ahead of the conference, Chief Technology Officer Kevin Scott told reporters and analysts the company is focused on helping spur the adoption of standards across the technology industry that will let agents from different makers collaborate. Agents are AI systems that can accomplish specific tasks, such as fixing a software bug, on their own. Scott said that Microsoft is backing a technology called Model Context Protocol (MCP), an open-source protocol introduced by Google-backed Anthropic. Scott said MCP has the potential to create an "agentic web" similar to the way hypertext protocols that helped spread the internet in the 1990s. "It means that your imagination gets to drive what the agentic web becomes, not just a handful of companies that happen to see some of these problems first," Scott said. Scott also said that Microsoft is trying to help AI agents have better memories of things that users have asked them to do, noting that, so far, "most of what we're building feels very transactional." But making an AI agent's memory better costs a lot of money because it requires more computing power. Microsoft is focusing on a new approach called structured retrieval augmentation, where an agent extracts short bits of each turn in a conversation with a user, creating a roadmap to what was discussed. "This is a core part of how you train a biological brain - you don't brute force everything in your head every time you need to solve a particular problem," Scott said.

Microsoft Aims for Cross-Company AI Agent Collaboration Through Open Standards

Improving AI memory is expensive but crucial for more meaningful interactions. Microsoft is advocating industry-wide standards to enable artificial intelligence (AI) agents from different companies to work together and retain more effective memories of their interactions, Chief Technology Officer Kevin Scott said on Sunday, according to Reuters. The remarks come ahead of the company's annual Build developer conference, which opens on May 19 in Seattle. Also Read: Google One Surpasses 150 Million Subscribers Amid AI Push Speaking at Microsoft's headquarters in Redmond, Scott emphasised the company's support for the Model Context Protocol (MCP), an open-source standard originally introduced by Google-backed Anthropic. The protocol aims to facilitate seamless communication among AI agents developed by different organisations, much like how the hypertext transfer protocol (HTTP) helped lay the foundation for the internet in the 1990s. Scott described the potential of MCP to create an "agentic web" -- a new layer of the digital world where intelligent agents can autonomously perform tasks and collaborate across platforms. "It means that your imagination gets to drive what the agentic web becomes, not just a handful of companies that happen to see some of these problems first," he said. Scott reportedly said Microsoft is also trying to help improve how AI agents remember and build on past user interactions. Scott acknowledged that most current AI interactions remain largely transactional, lacking continuity and context. Enhancing memory capabilities in AI agents, however, poses technical and financial challenges due to the significant computing resources required. Also Read: AI: Nvidia AI Factories in Saudi Arabia, DataVolt-Supermicro Deal, AWS-Humain AI Zone and More But making an AI agent's memory better costs a lot of money because it requires more computing power, according to the report. To address this, Microsoft is developing a technique called structured retrieval augmentation. The method allows AI systems to extract and retain concise, relevant information from each interaction, creating a structured summary or "roadmap" of the conversation. This mirrors how biological brains form memories -- not by recalling everything at once, but by focusing on key data points. "This is a core part of how you train a biological brain - you don't brute force everything in your head every time you need to solve a particular problem," Scott reportedly said.

Microsoft wants AI 'agents' to work together and remember things

REDMOND, Washington (Reuters) -Microsoft envisions a future where any company's artificial intelligence agents can work together with agents from other firms and have better memories of their interactions, its chief technologist said on Sunday ahead of the company's annual software developer conference. Microsoft is holding its Build conference in Seattle on May 19, where analysts expect the company to unveil its latest tools for developers building AI systems. Speaking at Microsoft's headquarters in Redmond, Washington, ahead of the conference, Chief Technology Officer Kevin Scott told reporters and analysts the company is focused on helping spur the adoption of standards across the technology industry that will let agents from different makers collaborate. Agents are AI systems that can accomplish specific tasks, such as fixing a software bug, on their own. Scott said that Microsoft is backing a technology called Model Context Protocol (MCP), an open-source protocol introduced by Google-backed Anthropic. Scott said MCP has the potential to create an "agentic web" similar to the way hypertext protocols that helped spread the internet in the 1990s. "It means that your imagination gets to drive what the agentic web becomes, not just a handful of companies that happen to see some of these problems first," Scott said. Scott also said that Microsoft is trying to help AI agents have better memories of things that users have asked them to do, noting that, so far, "most of what we're building feels very transactional." But making an AI agent's memory better costs a lot of money because it requires more computing power. Microsoft is focusing on a new approach called structured retrieval augmentation, where an agent extracts short bits of each turn in a conversation with a user, creating a roadmap to what was discussed. "This is a core part of how you train a biological brain - you don't brute force everything in your head every time you need to solve a particular problem," Scott said. (Reporting by Stephen Nellis; Editing by Paul Simao)