SAN FRANCISCO – Microsoft is launching autonomous artificial intelligence (AI) agents, or virtual employees, capable of performing routine tasks such as handling customer queries and identifying prospects, in an effort to speed up adoption of AI by businesses.

The US tech titan is also giving its customers the ability to create their own AI agents, and is launching 10 ready-to-use robots that can fulfill a range of roles including supply chain management and customer service.

AI agent has become the new buzzword for major software companies investing billions in powerful AI models, which they say will shape the future of computing.

Microsoft has moved quickly in bringing generative AI technology to everyday users, largely using models from its US$13 billion (S$17 billion) partnership with OpenAI, the maker of ChatGPT.

The agents are intended to advance generative AI from ChatGPT-style chatbots, which require human prompts, to systems that can operate without human intervention.

In a blog post, Microsoft introduced 10 autonomous agents for its Copilot AI platform, designed to support sales, service, finance and supply chain teams.

These agents will be available in public preview starting in December, with full deployment planned through 2025.

“Think of agents as the new applications of an AI-powered world. Every organization will have a constellation of agents – ranging from simple prompt-response to full autonomy,” explained Jared Spataro, Microsoft’s chief marketing officer.

The company envisions these agents facilitating the transition from today’s traditional business routines to new work habits driven and increasingly implemented by AI.

Additionally, Microsoft announced the upcoming public preview of Copilot Studio, a tool that allows businesses to create and manage their own AI agents.

Other tech giants are also exploring the deployment of AI agents, with venture capitalists pouring money into start-ups offering their own take on the concept.

Salesforce Chief Executive Marc Benioff, whose company in September announced its own roster of AI agents, criticized Microsoft’s efforts.

“When you look at how Copilot has been delivered to customers, it’s disappointing. It just doesn’t work and doesn’t provide any level of accuracy,” Mr. Benioff said last week.

AI chip giant Nvidia, Google and Oracle have also announced plans for AI agents.

Although their powers remain limited for now, the development of autonomous agents capable of acting independently has raised concerns about potential risks.

Microsoft has ensured that these new agents, currently limited to non-complex tasks, follow strict guidelines for security, privacy and responsible AI.

“Once these agents are created, IT administrators can apply a comprehensive set of features to govern their use,” he says. AFP