Forrester predicts that by decade's end, the rush toward agentic AI will grow so chaotic that CIOs will be forced into a new role as enforcer of order.

In a recent research note, the analyst warned that the promise of line-of-business departments building and deploying their own AI agents will fade as agent systems sprawl across the organization, increasingly misaligned with business needs.

Forrester said CIOS would end up "governing the enterprise AI-powered operating system" rather than running the tech. This is because the proliferation of AI agent systems, built into application software and cloud infrastructure, could lead to “fragmented adoption, weak data foundations, unclear decision-rights, or incomplete process design."

"In 2030, these errors will create systematic failure at scale. The challenge is no longer making transformation stick but ensuring the enterprise doesn’t outrun the capacity for control,” the research paper said.

This is the same Forrester which, in October last year, said fewer than one-third of decision-makers were able to make the connection between the value of AI and their corporation's financial growth. It meant large organizations were set to defer a quarter of planned AI spending from 2026 until 2027.

Enterprise application vendors are using their entrenched positions among customers to end discounting and push high-margin AI products, Forrester also said.

Nonetheless, AI agents will, in the end, come to dominate enterprise IT, it now claims.

• AI agents can't teach themselves new tricks – only people can
• Google claims to have all the answers for enterprise AI agent sprawl
• Google unleashes Gemini AI agents on the dark web
• AI agents are 'gullible' and easy to turn into your minions

"As software generates software and autonomous agents execute work, the CIO’s center of gravity shifts from building systems to governing outcomes," the latest paper said.

As such, CIOs must don a new cape, change their hats and find their way in the world wearing an entirely new costume fit for roles. Characters they may want to inhabit include: the architect of enterprise decision-making.

"Platforms must support real-time decision-making while enforcing constraints that protect margin and reduce downside risk," the paper said.

Another part will be the governor of autonomous systems. "CIOs must define bounded autonomy patterns that control how far an agent’s decisions propagate and how quickly the enterprise can intervene."

The more ambitious among CIOs might be tempted to play the role of "storytellers who translate probabilistic risk into confidence." That means, showing how uncertainty is "managed, monitored and owned," rather than brushed aside, the paper said. ®

More about

• AI