450 billion through AI agents and humans

According to a recent study by the Capgemini Research Institute, agent-based AI systems have an economic potential of up to 450 billion US dollars by 2028. This growth is primarily based on increased sales and efficiency gains. The greatest added value is created when people remain actively involved.

Autonomous AI agents

Trust in fully autonomous AI agents has fallen from 43% to 27% within a year - due to data protection and ethical concerns. Nevertheless, companies expect AI agents to play a role in most business processes over the next three years. Managers expect a 65% increase in value-adding tasks, a 53% increase in creativity and a 49% improvement in employee satisfaction with successful human-AI collaboration.

So far, only 2% of companies have fully scaled AI agents. Just under a quarter are testing in pilot projects, while 14% are in the implementation phase. 93 percent of decision-makers see scaling within twelve months as a competitive advantage - but almost half lack an implementation strategy.

Data maturity is crucial

The expected added value depends heavily on the degree of maturity: companies with scaled implementations could gain an average of 382 million US dollars over the next three years, while others could only gain around 76 million US dollars. In the short term, AI agents will primarily be used in customer service, IT and sales, while in the medium term they will also be used in operations, research and development and marketing. Challenges exist in terms of infrastructure and data maturity: 80 percent do not have a mature AI architecture, and less than 20 percent have a high level of data maturity. Data protection (51%), algorithmic distortions and a lack of explainability are additional obstacles - often without sufficient countermeasures.

"The potential of AI agents is enormous. The decisive factors are strategic realignment, ethics and safety as well as an organization that promotes human-AI collaboration," emphasizes Philipp Wagner, Vice President at Capgemini Invent Germany. "The key to success lies in a clear focus on results and redesigning processes with an AI-first mindset. Trust in AI comes from responsible development - with ethics and security as cornerstones. At the same time, companies must be organized in a way that enables effective human-AI collaboration - with the aim of strengthening human judgment and achieving better business results."