By 2025, artificial intelligence (AI)-powered agents are expected to work together to solve tasks in a collaborative system known as “multi-agent AI,” as predicted by technology services giant Capgemini.

This system involves a network of AI agents collaborating to solve tasks in a distributed manner, according to Capgemini.

Pascal Brier, Capgemini’s chief innovation officer, mentioned that companies are already exploring these agent technologies. He highlighted that we can expect to see applications using multiple autonomous agents as early as next year.

Capgemini describes AI agents as technology designed to function independently, capable of planning, reflecting, pursuing higher-level goals, and executing complex workflows with minimal or limited direct human oversight. Essentially, these AI agents work behind the scenes to complete tasks on your behalf.

Brier noted that the U.S. is further along in developing this technology, while Europe is somewhat behind.

In a recent research report titled “Harnessing the Value of Generative AI,” Capgemini found that the vast majority of companies surveyed (82%) plan to integrate AI agents within one to three years, while only 7% have no such plans. The report is based on a survey of over 1,100 companies with revenues of $1 billion or more.

According to Brier, there are two types of AI agents: individual agents that carry out tasks on your behalf and multi-agent technology, where agents communicate with each other. For example, a marketing AI agent could autonomously work with a legal department’s AI agent to ensure an ad campaign complies with regulations.

Unlike traditional AI systems that simply follow instructions, these new AI agents can understand, interpret, adapt, and act independently, potentially replacing human workers for certain tasks, Capgemini explained.

Brier recalled the first major wave of AI in 2022, which he refers to as “V1,” focused on understanding prompts and large language models (LLMs). Now, AI and generative AI are converging to build knowledge engines, using generative AI to interact with these engines, and employing AI agents as substitutes or co-pilots to accomplish tasks.

Capgemini’s research indicates that 71% of organizations expect AI agents to facilitate automation, while 64% believe they will free human workers from repetitive tasks, allowing them to focus on more valuable functions like customer experience.

In their report, Capgemini noted a fourfold increase in the number of organizations integrating generative AI into their operations. In 2023, 6% of firms were using generative AI, but this year, that number has risen to 24%. However, smaller companies are not adopting generative AI at the same rate as larger ones.

For example, only 10% of firms with annual revenues of $1 billion to $5 billion are implementing generative AI, while 49% of companies with annual revenues of $20 billion or more are doing so.

“The scale at which bigger companies are conducting generative AI experiments is larger, giving them more opportunities to measure results and move faster. They have also invested more than smaller companies,” Brier explained.

The adoption of generative AI also varies by industry. In aerospace and defense, 88% of organizations have invested in generative AI, while in retail, the figure is 66%.