Agenttic AI (systems that can plan, act, and learn on their own) are being adopted by organizations faster than traditional generative artificial intelligence. Within two years, 35% of companies are already considering implementing agent AI, and a further 44% plan to do so soon. However, few organizations have developed the management frameworks needed to redesign their workflows, governance models, investment plans, and talent strategies to keep up with this unprecedented pace.
These are some of the findings from the 9th annual global research study on AI and business strategy, released today by MIT Sloan Management Review and Boston Consulting Group (BCG). The report, titled “Emerging Agent Enterprises: How Leaders Should Navigate the New Era of AI,” is based on a survey of 2,102 executives and interviews with senior leaders across 21 industries and 116 countries.
“Historically, we have had a neat separation between technology and humans, and management processes have been designed around that distinction. But agent AI is neither a tool nor a teammate; it is both, and it thrives in that ambiguity. Successful organizations are those that recognize the duality of agent AI as a feature, not a bug,” said Sam Landsbotham, co-author of the report and professor of analytics at Boston University.
Unlike previous technologies, agent AI systems are more than just tools to be manipulated or assistants waiting for instructions. They act increasingly like autonomous teammates, able to execute multi-step processes and adapt as they go. Notably, 76% of executives surveyed view agent AI more like a colleague than a tool.
“Agent AI has the power to transform entire workflows and challenge existing business processes. Successful organizations are those that strive to reimagine their processes, rather than simply forcing agent AI onto existing processes,” said BCG Managing Director and Senior Partner Shervin Khodabande, leader of the firm’s AI business and co-author of the report.
Other key findings include:
- 66% of leading agent AI organizations anticipate changes to their operating model, compared to 42% of organizations just starting to use the technology.
- Fifty-eight percent of agentic AI leaders expect their governance structure to change within three years, and decision-making authority for AI systems is expected to increase by 250 percent.
- 43% of agentic AI leaders expect to be more open or willing to hire generalists instead of specialists. 45% expect to reduce their middle management ranks. 29% expect fewer entry-level roles.
- 95% of employees at leading agent AI organizations report that AI has had a positive impact on their job satisfaction.
- 73% of agentic AI leaders expect using this technology to change their organization’s ability to differentiate, compared to 53% of leaders without agential AI. Additionally, 76% of individuals working in leading agential AI organizations believe that the use of agentic AI will impact their ability to differentiate themselves from their peers, compared to 49% of individuals working in organizations without agential AI.
Due to the uniqueness of AI as both a tool, a colleague, and an agent, organizations are now faced with the unprecedented challenge of managing a single system that requires both an HR approach and asset management technology.
