In a recent DataIQ peer discussion, senior leaders across financial services, telecoms, insurance and the public sector compared notes on what happens after the first wave of agentic AI use cases. The pattern is familiar: strong executive interest, a growing portfolio of copilots and automations, but limited progress in embedding AI into how work is actually delivered.
What’s becoming clear is that scaling agentic AI is less about generating new ideas, and more about rethinking the structures that surround them.
DataIQ clients can find the full set of learnings here.
Click here to become a DataIQ client and access all the learnings, plus regular masterclass sessions and other peer-led discussions.
The shift from use cases to operating models
Most organisations are still operating within models designed for earlier generations of data and analytics. That creates friction when AI starts to move beyond isolated deployments.
Leaders described a common tension. On one side, business units continue to develop use cases. On the other, senior stakeholders are asking for enterprise-wide impact. Bridging that gap requires more than additional investment. It requires changes to how work is organised, owned and governed.
Rather than large-scale transformation programmes, the approach emerging in practice is more incremental. Organisations are making targeted adjustments across operating models, team structures and governance, often in parallel.
Where progress is being made
Three areas are beginning to act as focal points for change:
1. Reducing the friction between pilot and production
The challenge is not the lack of ideas but the complexity of moving them forward. Multiple approval pathways, unclear ownership and inconsistent value definitions can slow progress even where the use case is well understood.
As a result, attention is shifting towards simplifying how initiatives move through the organisation, rather than simply increasing their number.
2. Using domains to make change tangible
Rather than attempting organisation-wide transformation in one move, many teams are concentrating efforts within specific business areas.
By clustering related AI initiatives within a single domain, organisations can start to rethink how work flows end-to-end — not just optimise individual steps. This creates a more practical route from experimentation to something that resembles scale.
3. Rethinking roles as AI becomes embedded
As AI moves into workflows, it begins to change the nature of work itself.
Leaders are starting to question how roles will evolve, particularly where tasks can be automated or augmented. In many cases, the shift is gradual, with responsibilities moving towards oversight, judgement and exception handling. But the direction of travel is becoming clearer.
A broader organisational shift
These changes are not confined to data or technology teams. As AI starts to influence workflows and workforce design, ownership becomes more distributed.
HR functions are becoming more involved, particularly where the impact on roles and team structures is more explicit. At the same time, coordination across data, technology and the business is becoming more important, both to avoid duplication and to ensure that underlying data foundations can support more advanced use.
There is also a growing recognition that expectations around value need to be managed carefully. While early gains are often framed in terms of efficiency, the more meaningful impact is likely to take longer to materialise.
What this means in practice
For most organisations, the immediate challenge is not whether to invest in agentic AI, but how to create the conditions for it to scale.
That means looking beyond individual use cases and asking more fundamental questions about how work is structured, how decisions are made, and how different parts of the organisation coordinate around AI.
The shift is already underway. But it is uneven, and in many cases still being worked through in real time.
Become a DataIQ client to access the full set of learnings and many more exclusive insight pieces.
