Microsoft working on an OpenClaw-like agentic AI service to autonomously execute tasks in 365 Copilot

Microsoft is developing agentic artificial intelligence (AI) capabilities similar to those offered by OpenClaw for your service Microsoft 365 Copilot, in order to introduce AI tools that work autonomously 24 hours a day for Office applications, in a secure environment for companies.

The artificial intelligence (AI) agent OpenClawacts as an advanced personal assistant integrated into a computer, with the ability to control your computer locally to perform a wide variety of tasks such as managing files, emails or browsing the internet and applications.

However, since it became popular at the beginning of this year, there have been warnings about how the possibility of accessing the entire computer can pose a risk to cyber attacks or extensions that conceal ‘malware’, among other security-related issues.

In this sense, Microsoft this “exploring the potential of technologies like OpenClaw in an enterprise context“, working on the development of autonomous AI agents that run locally, avoiding security concerns that comes with the use of OpenClaw.

This was stated by Microsoft’s corporate vice president, Omar Shahine, in statements to The Information collected by The Verge, where he indicated that Microsoft is working on a technology with “safer” versions of the AI ​​tool for its 365 Copilot service, with the idea that “operate autonomously 24 hours a day” performing tasks on behalf of users.

Specifically, it is expected that this technology in Microsoft 365 Copilot can perform tasks such as continuously monitoring your inbox and calendar of a user’s Outlook, as well as displaying task suggestions.

Likewise, the aforementioned media has indicated that, beyond working with Office applications, the technology company is also exploring the ability to use autonomous agent AI to other tasks related to marketing, sales or accounting.

In this case, it would allow agents to be used for those specific tasks limiting your permissions to your specific activity and, with this, isolating their abilities to access other business environments, which would guarantee more security.

With all this, it is expected that Microsoft Introduce these new agentic AI tools during their developer conference Microsoft Buildscheduled to be held on June 2 and 3 in San Francisco (United States).

It is worth remembering that Microsoft has been working on the incorporation of agentic AI services for companies for some time, for example, with the recent incorporation of Claude Cowork of Anthropic and Microsoft 365 Copilot or the general availability of Agent 365 and Microsoft 365 E7 Frontier Suite.

Likewise, it also announced the evolution of Microsoft 365 Copilot from assistant to agent integrated with Copilot Cowork, powered by ‘Work IQ’, an AI layer that allows Microsoft 365 Copilot and agents to know more information about the job itself and its environment.