As organisations shift from isolated pilots to enterprise-wide deployments of generative and agentic AI, they are unlocking transformative benefits in innovation and productivity.

But mainstream adoption is bringing new challenges related to cost containment, workforce adaptation, governance, and sustainability.

Harnessing the value of AI: Unlocking scalable advantage, the third edition in the Capgemini Research Institute’s annual research series on AI technologies, explores strategies for how organisations can scale AI implementation responsibly, ethically, and effectively. The research brief is based on findings from a global survey of 1,100 leaders at organisations with annual revenue above $1 billion across 15 countries. Key findings include:

  • Gen AI adoption is now mainstream, surging from 6% in 2023 to 30% in 2025. Today, 93% of organisations are exploring or enabling Gen AI capabilities – yet while benefits are rising, cost concerns persist.
  • AI agents are gaining ground, with 14% of organisations implementing them at partial or full scale, and 23% running pilots. Of the organisations already scaling AI agents, nearly 45% are piloting or scaling multi-agent systems.
  • AI is evolving from tool to teammate. Nearly six in 10 organisations are planning to integrate AI as augmenting or autonomous collaborators within the next year – yet most are underprepared for this shift.
  • Trust and governance are lagging: 71% of organisations say they cannot fully trust autonomous AI agents for enterprise use. While 46% have governance policies in place, adherence remains low.
  • AI’s environmental impact is under scrutiny. Only one in five organisations measures its Gen AI environmental footprint, though sustainability measures – like using smaller task-specific models – are gaining traction.

The new research brief offers actionable insights for business and technology leaders across industries and functions. To deliver business value and scale AI responsibly and effectively, organisations must:

  • Architect for scalability by redesigning processes for AI integration, and embracing “platformisation” for enterprise-wide deployment.
  • Reinforce trust through governance by defining clear scopes for AI execution, establishing cross-functional governance with ethical oversight, and strengthening data management and traceability.
  • Design human-AI collaboration models through prioritising reskilling and cultural transformation, and adapting workflows and performance metrics for hybrid teams.

To discover how organisations can move beyond experimentation to scaled, ethical, and high-value AI deployment, download the Harnessing the value of AI research brief today.