Starting in November 2024, Microsoft will enable its customers to create autonomous AI agents, a development aimed at improving its stake in the artificial intelligence sector.
These autonomous agents, which will function with minimal human oversight, are termed “apps for an AI-driven world” and are expected to automate key business processes such as customer service, sales lead identification, and inventory management.
Microsoft’s new offering will be accessible through its Copilot Studio application, which is designed to be user-friendly, requiring little technical expertise to create and deploy the AI agents. The agents will leverage a mix of proprietary AI models and those developed by OpenAI, increasing the scope of possibilities for businesses eager to simplify operations.
As part of its initiative, Microsoft will also roll out 10 pre-built AI agents which will handle common business tasks, such as managing supply chains, tracking expenses, and enhancing client communications.
McKinsey & Co., a consulting firm with early access to these tools, showcased how an AI agent could effectively manage client inquiries, track interaction history, assign tasks to the right consultant, and schedule follow-up meetings.
The technology will bolster businesses, as Microsoft envisions a future where every employee has a personal AI assistant, or “Copilot,” to interact with the various autonomous agents within their organisation.
Charles Lamanna, corporate vice president of business and industry Copilot at Microsoft, explained that the Copilot will serve as a personal interface for employees, providing them with a tailored experience while interacting with an array of AI tools to boost productivity.
This latest development comes despite a slight dip in Microsoft’s stock performance for the September quarter, the company is focused on the long-term prospects of its AI technologies.
Microsoft’s announcement follows Meta’s unveiling of its “Self-Taught Evaluator” model, which focuses on creating AI that can train and evaluate itself with minimal human involvement. More autonomous systems capable of self-learning are being built, potentially reducing reliance on human-generated data and expert intervention.
