Microsoft is building an OpenClaw-like AI agent for Copilot with enterprise security in mind

Microsoft is working on yet another OpenClaw-like agent | TechCrunch

Microsoft is testing ways to integrate OpenClaw-like features into its existing Microsoft 365 Copilot tool. The new features, which the company confirmed to The Information, would be geared towards enterprise customers, with better security controls than the famously risky open source OpenClaw agent. OpenClaw is a tool that runs locally on a user's computer and can create agents to perform tasks on behalf of the person. If Microsoft does come up with its own version of a Claw -- meaning an agent that runs locally -- the effort would join a number of other agentic tools the company has announced in the past few months. In March, for instance, Microsoft announced Copilot Cowork, which is designed to take actions in M365 apps, not just provide search results or chat in a separate work pane. Cowork is powered by its own "WorkIQ" technology, an intelligence layer that tries to personalize Cowork for the user across Microsoft 365 apps. Microsoft has also tapped Anthropic Claude's to power Cowork, after it partnered with the AI lab late last year. Microsoft added Claude as an option available for Cowork. (While OpenClaw can work with multiple models, Claude remains the model of choice for many users of the open source project.) However, Cowork doesn't run on the local hardware; it runs in the cloud. In February, Microsoft also introduced Copilot Tasks, another agent designed to complete tasks, which was released at the time in preview. The marketing materials made this agent sound like it was geared more towards prosumers than enterprises, with tasks it should be allowed to handle ranging from organizing email (a M365-like task) to organizing travel and appointments (tasks outside Microsoft's Office suite of apps). This, however, also runs in the cloud. It's not clear yet whether this Claw would be local or if it would simply adopt some of the other features that OpenClaw advocates love. Microsoft did tell the Information that one of the main features of the agent is that it would essentially be a version of 365 Copilot that is always working, able to take actions at any time. The idea is an agent that can complete multi-step tasks over long periods. While the open source project OpenClaw can run on Windows machines, the Mac Mini has been the go-to platform for OpenClaw users. So much so, that the small, affordable, cube-shaped Mac Mini desktops have suddenly been selling like hotcakes. So, beyond security, Microsoft may have a number of motivations for wanting its own version. The company is expected to show off this new Claw (or an upgraded version of one of its existing Claw-like tools) at its Microsoft Build conference in June, The Verge reports. We've asked Microsoft how the new Claw agent fits in with these other agents and will update the story when we receive comment.

Microsoft Plans to Bring Copilot Into the Agentic AI Age

Blake has over a decade of experience writing for the web, with a focus on mobile phones, where he covered the smartphone boom of the 2010s and the broader tech scene. When he's not in front of a keyboard, you'll most likely find him playing video games, watching horror flicks, or hunting down a good churro. Microsoft is leaning into agentic AI with a planned revamp of its own AI assistant, Copilot, according to a new report from The Information. The always-on Copilot, when launched, will be able to complete tasks for you, not unlike the viral OpenClaw platform. OpenClaw, the open-source platform for creating AI agents, catapulted agentic AI into the spotlight ever since it was released, and major AI players are jumping in headfirst. Nvidia recently released its own reference stack, NemoClaw, which provides some much-needed safety guardrails that OpenClaw doesn't currently have, such as the ability to track all actions the AI agent takes. This was shortly followed by Anthropic's announcement that subscribers with certain plans can have Claude perform tasks on their behalf. OpenClaw's popularity seems to have piqued Microsoft's interest, too. Omar Shahine, Microsoft's corporate vice president, told The Information that the tech giant is looking into OpenClaw-like technologies. According to the report, sources also said Microsoft is working to make these technologies safer. Safety is a concern with OpenClaw. There are virtually no security or privacy measures in place, making the platform feel like the wild west. It's why the likes of NemoClaw exist, so Microsoft prioritizing safety comes as no surprise if it's expected to deploy the new Copilot for enterprise. An agentic version of Copilot may be able to dip into your email and calendar to generate a to-do list for you each day. This would likely be just the beginning of the tool's functions, with more features to follow and further integration into Microsoft products. We might not have much longer to wait before we see the new agentic Copilot, either. Microsoft's developer conference, Build, is set to take place June 2-3, where AI will undoubtedly be a primary focus. If Microsoft can succeed in producing a safer agentic assistant with Copilot, it might steal some of OpenClaw's attention. Microsoft representatives didn't immediately respond to a request for comment.

Microsoft is testing OpenClaw-like AI bots for 365 Copilot

Microsoft is looking into ways it can integrate OpenClaw-style features into 365 Copilot, according to a report from The Information. The test reportedly comes as part of efforts to make its 365 Copilot AI assistant "run autonomously around the clock" while completing tasks on behalf of users. Omar Shahine, Microsoft's corporate vice president, confirmed to The Information that the company is "exploring the potential of technologies like OpenClaw in an enterprise context." OpenClaw is an open-source platform that allows users to create AI-powered agents that run locally on a user's device. The platform rose in popularity earlier this year, though it has since raised a number of serious security concerns. Sources tell The Information that Microsoft is confident that it can implement "safer" versions of the tool. The always-on version of 365 Copilot could reportedly do things like monitor a user's Outlook inbox and calendar and serve up a list of suggested tasks each day. Microsoft is also exploring OpenClaw-like agents tailored to certain roles, such as marketing, sales, and accounting, to "limit the permissions the agent needs," siloing them from other parts of a business, according to The Information. As noted by The Information, Microsoft aims to show off some of these features during its Build conference, which kicks off on June 2nd. Last year, Anthropic launched integrations with its Claude AI chatbot inside Microsoft 365 services, while bringing its Claude Cowork tool to Copilot to help complete "long-running, multi-step tasks." Bringing OpenClaw-like capabilities into Copilot could help Microsoft reclaim some of the customers it lost to rival services.

Microsoft wants to build its own, more secure version of OpenClaw for Copilot

* Microsoft is reportedly working on an OpenClaw alternative * Copilot users are on the up, but nowhere near ChatGPT levels of engagement * New M365 E7 and Agent 365 plans target increased demand for AI agent management Microsoft is thought to be working on its own autonomous AI agents, inspired by OpenClaw. Per The Information reporting, the proposed agentic platform could replace or work alongside existing Copilot tools, bringing proactive, always-on style AI to workers who, if they choose Microsoft's AI today, only have reactive assistants. It's unclear whether the company would target local or cloud-based agents, or a hybrid of the two, but if today's Copilot is anything to go off, they would likely either only work within Microsoft's ecosystem or, at best, work best there. Microsoft wants to build proactive AI agents If Microsoft were to come out with its own OpenClaw-like agentic platform, it would likely target enterprise customers at first, building on existing Copilot Cowork development. We could see role-specific agents for sectors like sales, marketing and accounting, and likely access controls and permissions on a per-agent basis, with AI agents being increasingly treated like humans. The Information's sources also claim Microsoft is targeting agents that are "safer" than competition (read OpenClaw). Compared with other hyperscalers and AI companies, Microsoft initially fell short on the AI front. ChatGPT is still, by far, the most popular AI chatbot, accounting for four in five sessions globally (per Statcounter). Maybe Microsoft hopes that getting ahead of the curve on the proactive agentic front could regain lost enterprise customers. The company recently confirmed the launch of a new E7 tier, which bundles parts of Microsoft 365, Copilot and Agent 365 for a new world in which bosses are expected to manage humans and AI agents simultaneously. Microsoft also recently declared Copilot users had grown 160% year-over-year, and daily active users were up ten times. Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button! And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.

Microsoft to Add OpenClaw AI Agent for Smarter Microsoft 365 Copilot Tasks

The Microsoft OpenClaw AI agent may work smartly by using both cloud and local systems. Local use can ensure the safety of important information and also speed up the system. Security is the highlight of this update. to add strict controls, limited access, and proper tracking. This will help companies feel safe while using AI agents for important workflows. Many companies now demand AI agents that can work without constant input. These tools save time by handling simple and repetitive tasks. The is built to meet this need. Other tech companies are also working on AI agents. Microsoft wants to stay strong in this race by offering a safe and useful solution for businesses. More details may come at the Microsoft Build event in June. The company may show how this new AI agent works or improve its current Copilot features. This shows how AI is changing office work. are not just for answering questions now. They can do real tasks and help people finish work faster. Microsoft is trying to make this change simple and safe for everyone.

Microsoft reportedly building OpenClaw-style AI system

New AI agents may run on devices and handle daily work, but raise safety concerns. Microsoft is reportedly working to improve its AI tools so they can make them more proactive and useful in daily work. Instead of waiting for users to ask questions, the AI tool could start doing tasks in the background. As per the current reports, the company is testing new features for its Copilot assistant that work a bit like smart helpers. These helpers could handle everyday work without needing constant instructions. This change could make AI tools more useful and save time. But it also brings worries about safety and how much control users will have over what the AI does. Microsoft says it is focusing on building a secure and enterprise-ready version of this approach. Microsoft wants to make Microsoft 365 Copilot more independent. It is trying out new functionalities that have been introduced by OpenClaw. The software will allow users to run AI agents right from their computers. The AI agents will perform the assigned tasks independently. While some individuals are optimistic about the innovation, others are concerned about possible security threats. Also read: One UI 8.5 release date for Samsung Galaxy S25 Ultra, S25+, S25 leaked, launch may be sooner than expected Omar Shahine, a senior executive at Microsoft, confirmed that the company is studying how such technology could fit into enterprise use. The goal is to create a safer version that businesses can trust. One idea under consideration is an always-on Copilot that monitors tools like email and calendars. It could then suggest daily tasks or reminders without being asked. Microsoft is also experimenting with role-specific AI bots that can be developed specifically to work for certain teams, like marketing, sales, and accounting. With the help of the availability of data for each bot, Microsoft seeks to reduce potential risks and safeguard confidential data. Such a step may prove useful when it comes to dealing with problems arising after the growing popularity of OpenClaw. Also read: OpenAI touts Amazon alliance in leaked memo, flags limits of Microsoft deal The company is expected to reveal more details at its upcoming Microsoft Build conference, which begins on June 2. This event often serves as a platform for major announcements related to AI and software development. Last year, Anthropic introduced integrations of its Claude AI into Microsoft's ecosystem, and expanding Copilot's abilities could help Microsoft stay competitive and win back users who have explored other AI services.