Four tensions of the agentic company
That those who are ahead gain an advantage is expected, and it is what gets quoted from the report. What I take from it are the four tensions, because none of them is solved by buying software: they are decisions about who has authority over what.
The 2025 report by MIT Sloan Management Review and Boston Consulting Group reveals that organizations adopting agentic artificial intelligence are gaining a competitive advantage. However, this implementation creates four distinct tensions that leaders must learn to manage.
Imagine a supermarket chain that installs an agentic AI system able to reorder inventory and set prices without human intervention. Its sales rise 12% in three months, but at the same time its floor managers start to question their own role. That paradox is what many organizations are living through as AI moves from being a tool to being an operational agent inside the company.
The 2025 report by MIT Sloan Management Review and Boston Consulting Group, published in MIT Sloan Management Review, was based on interviews with executives and performance data from hundreds of companies that had already implemented agentic AI. The authors compared growth, profitability and staff turnover indicators before and after adoption, and cross-referenced those results with surveys on perceived control and strategic alignment.
Four tensions emerged from that analysis and appear repeatedly: the loss of control over critical decisions and the need to redefine responsibility when AI acts autonomously; the pressure on organizational culture when hybrid roles between humans and machines are introduced; the management of bias risks and regulatory compliance in automated decisions; and the resistance of talent who see their work threatened by automation. Each of them translates into hidden costs, such as higher turnover (up to an additional 8%) or training investments that can absorb up to 15% of the innovation budget.
Leaders who want to capitalize on the competitive advantage need to create governance structures that clearly delimit which decisions can be delegated to AI and which require human oversight, establish algorithm audit protocols and dedicate resources to reskilling their teams. Only by aligning the technology with a talent management and ethics strategy can the advantage be kept from turning into a source of internal friction.
Conclusions
- Companies implementing agentic AI are already gaining a competitive advantage.
- Adopting this kind of AI introduces four types of tensions that leaders must anticipate and handle.
- Managing these tensions effectively is essential to fully capitalize on the benefits of agentic AI.
- Leaders' ability to navigate these complexities will define success in the new era of AI.