Rising value of AI is leading to decreasing trust in its reliability
The latest research report, titled "Rise of agentic AI: How trust is the key to human-AI collaboration," has revealed a significant shift in the adoption of AI agents by organizations worldwide. The survey involved participants from the USA, UK, France, Germany, and India, with preferred partners for implementing agentic AI systems including Microsoft, IBM, and Google.
The report suggests that most organizations are still in the early stages of implementing AI agents. Only 2% have implemented AI agents at scale, while 12% have achieved partial-scale implementation. A vast majority (over 80%) of organizations lack a mature AI infrastructure.
Despite the slow pace of full-scale implementation, the business functions expected to see the greatest adoption of AI agents are customer service, IT, and sales. Organizations prefer to use AI agents that are already available within existing enterprise solutions, with 62% partnering with solution providers like SAP or Salesforce.
The report also highlights the importance of human-agent collaboration in building trust. Integrating human involvement with processes handled by AI agents delivers a 65% increase in employee engagement on high-value tasks and a 53% increase in creativity. This collaborative approach is expected to become the prevailing model within one to three years, with AI agents acting as integrated members within human-supervised teams.
However, nearly two in five executives believe the risks of implementing AI agents may outweigh the potential benefits. The share of organizations expressing trust in fully autonomous AI agents has decreased from 43% to 27% in the past year. Fewer than one in five organizations report having high levels of data readiness, and only half of the organizations surveyed report having adequate knowledge and understanding of AI agents and their capabilities.
The total economic value of agentic AI systems is projected to reach $450 billion by 2028 in surveyed countries. Most organizations envision AI agents augmenting human team members in the next 12 months. A smaller but significant group, 53%, opts for a hybrid approach that combines in-house development with external solutions.
The majority of organizations are much further behind: 23% are in the process of piloting some use cases, 30% have just started exploring the potential, and another 31% are only considering experimenting within the next six to twelve months. Within one to three years, the percentage of processes handled by AI agents is projected to increase from 15% to 25%.
In conclusion, the report underscores the need for organizations to focus on building trust in AI agents through human-agent collaboration. As the adoption of AI agents continues to grow, understanding their potential benefits and risks, and fostering collaboration between humans and AI, will be key to their successful implementation.
