Executives are being pushed to act quickly on artificial intelligence, but the harder question is not how fast to buy it. It is whether the business is ready to absorb it. Across enterprises, the evidence increasingly points to the same conclusion: AI does not repair weak operating models. It amplifies them.
That is why the most effective organisations are shifting their attention from model selection to process readiness. The logic is straightforward. If data is inconsistent, ...
Continue Reading This Article
Enjoy this article as well as all of our content, including reports, news, tips and more.
By registering or signing into your SRM Today account, you agree to SRM Today's Terms of Use and consent to the processing of your personal information as described in our Privacy Policy.
workflows are fragmented and governance is vague, AI will not quietly smooth out the problems. It will expose them faster, and often at greater cost.
Gartner warned last year that at least 30% of generative AI projects would be abandoned after the proof-of-concept stage by the end of 2025, largely because of poor data quality, weak controls, rising costs and a lack of clear business value. The pattern is echoed elsewhere. A Boomi report cited by TechRadar found that while 86% of organisations had moved AI into production, only 34% trusted what their AI agents were doing. The issue was not model capability, but the state of the underlying data, integration and governance.
That mistrust is costly. If an automated recommendation is built on stale records, duplicated entries or conflicting systems of record, confidence in the technology can collapse quickly. Once staff see AI producing polished but unreliable outputs, scepticism tends to spread beyond one use case and into the wider transformation programme.
The same problem appears in agentic AI, where autonomy is being added to systems that are not always ready for it. ITPro reported that about half of agentic AI projects are still stuck in proof of concept, even as many enterprises continue to increase budgets. Gartner has separately warned that a sizeable share of organisations may demote or shut down AI agents within the next year because governance has not kept pace with deployment. In practice, firms are discovering that access rights, oversight and accountability cannot be treated as an afterthought.
There is also a cultural dimension. TechRadar recently reported that AI momentum in the UK is stalling in part because employees do not always trust the systems or feel comfortable with the changes they bring. In many organisations, enthusiasm fades after the first round of experiments, and teams drift back to familiar manual processes. Technology alone cannot solve that either. If people are not aligned to the change, the tools will not deliver the promised shift in performance.
This is why a process-first sequence matters. Before an enterprise introduces AI, it needs to know whether its operating model is stable enough to support it. That means assessing ERP performance, cleaning up data quality, clarifying ownership, standardising workflows and connecting systems that currently operate in isolation. Only then does it make sense to introduce automation that depends on trustworthy information.
Microsoft has made a similar point in its own positioning around agentic business applications, stressing that business transformation requires functional leaders to align processes with new capabilities, and that structured data, governance and business logic must come first. In other words, intelligence is only as useful as the environment it sits on top of.
For businesses under pressure to show progress, that may sound slower than simply launching an AI pilot. But it is often the faster route to a durable result. The alternative is to spend on licences, integration and change management only to discover that the organisation must revisit the same process issues later, after trust has already been weakened.
The more responsible approach is controlled autonomy: human oversight, defined boundaries and clear governance before machines are asked to make or influence decisions. That is particularly important in regulated sectors, where a system that cannot be explained after the fact can create compliance risk as well as operational inefficiency.
The businesses that get the most value from AI are unlikely to be those that moved first. They are more likely to be the ones that prepared properly, with connected systems, governed data and stable processes in place before asking AI to act.
Source: Noah Wire Services
