Brief IA

Microsoft and OpenClaw: Towards a Safer Agentic Copilot

💻 Code & Dev·Tom Levy·

Microsoft and OpenClaw: Towards a Safer Agentic Copilot

Microsoft and OpenClaw: Towards a Safer Agentic Copilot
Key Takeaways
1Microsoft plans to transform its Copilot assistant into an agentic AI tool, inspired by the OpenClaw platform.
2OpenClaw, while popular, lacks security measures, prompting Microsoft to prioritize this aspect in its redesign.
3Microsoft's Build conference in June could unveil advancements on this new agentic Copilot.
💡Why it mattersThe integration of a secure agentic Copilot could strengthen Microsoft's position against its competitors in the AI field.
Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisis ton rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

📄
Full Analysis

Microsoft is set to take a significant step in the field of agentic AI with a major overhaul of its AI assistant, Copilot. According to a report from The Information, this enhanced version of Copilot will be designed to perform tasks autonomously and will always be active, drawing inspiration from the model proposed by the OpenClaw platform.

OpenClaw, an open-source platform dedicated to creating AI agents, has recently captured the attention of major tech players. Since its launch, it has highlighted the potential of agentic AI. Nvidia, for example, has introduced NemoClaw, which includes essential safety guardrails, such as the ability to track all actions performed by the AI agent, a feature that OpenClaw currently does not offer. Anthropic has also announced that its assistant Claude could perform tasks for certain subscribers, underscoring the growing interest in these technologies.

Microsoft is not lagging behind. Omar Shahine, corporate vice president at Microsoft, indicated that the company is actively exploring technologies similar to OpenClaw. According to the report, Microsoft is working to make these technologies safer, a crucial aspect for their integration into professional environments.

A Microsoft spokesperson confirmed to CNET that the company is collaborating with OpenClaw to experiment with ways in which AI can facilitate users' work. "As customers adopt AI at scale, we continue to experiment with capabilities that move from conversation to action on your behalf," he specified. The spokesperson added that while the work with OpenClaw is still in its early stages, it reflects Microsoft's goal of responsibly learning to reduce daily friction, allowing users to focus on the most important tasks.

Security is a major concern for OpenClaw, often compared to a "wild west" due to the lack of protective measures. Microsoft intends to prioritize this aspect for its new Copilot, especially if it is intended for professional use. An agentic version of Copilot could access personal tools like emails and calendars to organize daily tasks, with additional features to come.

Microsoft's Build conference, scheduled for June 2 and 3, could be the opportunity to unveil these innovations. If Microsoft succeeds in developing a safer agentic assistant with Copilot, it could divert some attention away from OpenClaw. For now, Microsoft has not provided further comments, but the development of a secure agentic Copilot could well strengthen its position against competitors in the AI field.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.