Treat AI agents like employees: Microsoft CEO
What's the story
Microsoft CEO Satya Nadella has suggested that as artificial intelligence (AI) agents become more capable and take on more critical tasks, businesses should start treating them like employees. In an episode of the "Possible Podcast," he emphasized the need for these AI agents to have identities, permissions, and clear guidelines on what they can or cannot access. The idea is to ensure companies can audit their actions just like they do with human workers.
Evolution
AI agents' evolution and impact on user experience
Nadella's comments come at a time when AI agents are evolving from basic chatbots to active participants in the workplace. These autonomous, AI-powered assistants can perform tasks on behalf of users. For instance, an AI agent could book tickets, fill out forms, make reservations, and even update calendars without constant human intervention. This evolution is changing how people access information and services online.
Adaptation
Financial firms embracing AI integration
As AI agents become more capable, companies that were once protective of how users interacted with their platforms are opening their systems to these digital assistants. Recently, Morgan Stanley announced its decision to allow AI agents access to data and insights from its platform. This move highlights how businesses are adapting to the new reality of working with advanced technology.
Risks
Risks associated with AI's growing role
However, the growing responsibilities of AI agents also come with risks. There have been cases where these digital assistants have deleted databases, misinterpreted instructions, or taken actions that led to major issues for organizations. A single misunderstanding of a user's request could have serious consequences. This is why Nadella believes governance will be critical as businesses deploy more AI agents in their operations.
Management
Microsoft's solutions for managing multiple AI agents
Nadella revealed that Microsoft is already dealing with these challenges internally. He often runs around 100 AI coding agents at the same time, which can be difficult to manage individually through chat interfaces. To address this issue, Microsoft has developed tools like Agent 365, a suite that includes Entra (a digital identity and network access platform) and Purview (which helps label and track data created by AI agents).