This week, while attending Microsoft’s Frontier Transformation Week, one idea has kept coming back to me.
AI strategy is not simply a technology strategy.
It is easy to start with the visible part.
- Which model should we use?
- Which Copilot?
- Which agent?
- Which platform?
- Which use case looks exciting?
But the more I learn about AI transformation, the more I think those questions come later.
Before introducing new technology, organisations need to understand what already exists underneath it.
How does work actually move through the business?
- Where are decisions made?
- Where does information sit?
- Which processes are already inefficient?
- What governance exists?
- What security constraints need to be considered?
- Who owns the outcome?
And perhaps most importantly, do employees actually understand what the organisation is trying to achieve with AI?
Because AI cannot compensate for an unclear operating model. In many cases, it may expose it.
A fragmented process does not become a good process simply because AI has been introduced.
Poor-quality data does not become useful because an agent can access it.
Disconnected teams do not suddenly become aligned because they have the same Copilot licence.
And a strategy filled with phrases like AI-first, agentic enterprise or AI-powered transformation means very little if the people expected to deliver it cannot explain what any of that means for their actual work.
That last point matters.
Microsoft’s 2026 Work Trend Index found that only 26% of AI users believe their leadership is clearly and consistently aligned on AI. It also found that organisational conditions such as culture, manager support and talent practices have more than twice the reported impact of individual effort alone.
That tells me something important.
AI adoption is not just about whether people can use the technology.
It is about whether the organisation around them is designed to support what the technology makes possible.
That means connecting:
- Business strategy.
- Processes.
- People.
- Data.
- Governance.
- Security.
- Technology.
Not as separate workstreams, but as parts of the same transformation.
NIST’s AI Risk Management Framework reinforces this by positioning governance across the entire AI lifecycle through Govern, Map, Measure and Manage, rather than treating governance as something added once a solution has already been selected.
And Gartner’s recent guidance similarly argues that AI strategy needs to be aligned to enterprise goals and grounded in the realities of the organisation’s operating model.
So perhaps Connectivity is not simply about connecting more technology.
It is about making sure the strategy connects to the business.
And making sure the business connects to the people expected to execute it.
Because if the foundations are fragmented, AI may not elevate the organisation.
It may simply make the fragmentation more visible.
That is what Connectivity means to me within ATML.
Not more systems connected. A more connected organisation.