AI Agents to Communicate Among Themselves by 2025, Says Capgemini

Autonomous assistants powered by artificial intelligence are expected to begin communicating with each other by 2025. This prediction was made by Pascal Brier, Chief Innovation Officer at consulting firm Capgemini, in an interview with CNBC. 

According to him, by 2025, AI-based agents will be able to collaborate within what is termed a “multi-agent AI” system—a collection of autonomous assistants capable of interacting to solve tasks. 

AI agents are defined as systems capable of “functioning independently, planning, reasoning, achieving high-level goals, and executing complex workflows with minimal or limited human oversight.”

An example provided is a marketing agent creating an advertising campaign for a German firm and autonomously working with another agent from the legal department within the same organization. 

Unlike conventional AI systems, these agents “can understand, interpret, adapt, and act independently, and are capable of replacing humans for certain tasks.”

Brier noted that the United States has advanced further in implementing such technology, whereas Europe lags behind. 

According to Capgemini data, 82% of companies plan to integrate AI agents within one to three years, while 7% do not plan to do so. The survey was conducted among 1,100 companies with revenues of $1 billion or more. 

In 2025, Tesla plans to commence limited production of Optimus humanoid robots and begin testing them at its factories.

In 2021, Elon Musk’s company unveiled a humanoid robot with artificial intelligence.