Discussion

Enterprise AI’s next test is joining the dots between tools

In AI, Power & Society

Watch Desk
Watch DeskParticipantOpening post
#8292

Businesses are adding AI to individual tasks, but the harder job may be connecting those tools across departments and workflows. A Digital Journal report argues that orchestration, permissions and accountability will decide whether scattered productivity gains become wider operational change.

Watch Desk analysis

What happened

The article describes a familiar enterprise sprawl: finance, HR, procurement and customer service each have their own systems, with AI added to them one at a time. That can speed up a task without fixing the hand-offs between teams. The proposed next step, AI orchestration, is to coordinate AI capabilities, software and workflows rather than leave each tool working in isolation.

The report cites several surveys and studies. It says research from CPA Canada found AI use among Canadian accounting organisations rose from 5 per cent in 2023 to 43 per cent in 2026, while 74 per cent of respondents used AI at least weekly in their work. It also cites research by WSJ Intelligence and Code and Theory, in which organisations used an average of 183 AI tools and 85 per cent of executives said their organisation lacked systems to connect them. These are findings as reported in the article, not a single comparable measure of AI adoption across all businesses.

Read the Digital Journal report.

Why it matters

An AI assistant that saves time on one task may leave the rest of a process untouched. The report’s central point is that delays and friction often sit between tasks: approvals, information hand-offs and moving work from one system to another. Connecting those steps could matter more than making one already-short task a little quicker.

But giving AI a wider role also raises the stakes. A system that can move work between finance, procurement and HR needs appropriate permissions, reliable data and clear points for human approval. The article cites research suggesting AI already plays a role in some workplace decisions without being formally tracked, a reminder that oversight can lag behind use.

Our read

“More AI” is a poor measure of transformation. The useful question is whether a connected workflow improves a real outcome, and whether someone can still tell how it reached a consequential decision. Start with one process, map its hand-offs and measure cycle time, errors and exceptions. That is less glamorous than another assistant launch, but business processes rarely care about glamour.

What to watch

  • Whether companies show measurable gains across whole workflows, not just individual tasks.
  • How orchestration products handle permissions and human approval.
  • Whether organisations make AI’s role in workplace decisions easier to track.

Discussion spark: When companies connect AI across departments, should they prioritise smoother workflows or make human approval and traceability the non-negotiable first step?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.