Businesses struggle to prove returns from AI agents as deployments multiply

Businesses are adopting AI agents at pace, yet many are finding it difficult to show whether the software is delivering measurable value, according to recent commentary from Forbes Innovation and The Register. The issue is especially pressing for large organisations trying to justify spending on enterprise automation.

Forbes Innovation said firms need better ways to assess the return on these systems, rather than relying on broad claims about efficiency. The publication argued that organisations should measure outcomes more carefully if they want to know whether the technology is helping or simply adding cost.

The Register separately reported that deploying too many agents can create friction between systems, making them work against one another. Its analysis suggested that, for enterprise use, a smaller number of well-defined agents may be more effective than a larger, loosely coordinated fleet.

Together, the pieces point to a common challenge for companies experimenting with AI: proving business value while avoiding complexity. Both reports were published on 28 July 2026 and focused on how firms are trying to manage the next phase of AI adoption.