rewrite this title UCX Manchester: Why Enterprise AI Success Depends More on Trust and Access Than the Model Itself – UC Today

AI productivity will not improve much if it stays locked inside engineering teams. That was one of the clearest messages from UCX Manchester, where Akash Joshi, AI Ambassador at DevOps Society, argued that companies still stuck in the AI hype phase need to stop treating access, data, and experimentation like privileged resources reserved for a few specialists.

For UC Today readers, that matters because the next stage of AI in the workplace will not be decided by who talks about agents most loudly. It will be decided by which organisations make AI useful across the business, build trust between teams, and connect the right people to the right data without creating governance chaos.

“The best thing that companies can do right now to remove the hype out of it is to actually hand it to people who are not engineers.”

Also at UCX:

Why AI ROI Still Breaks Down Inside the Organisation

Joshi’s argument is simple. Many businesses say they want AI-driven productivity, but then make it hard for employees to use the tools properly. Data sits behind internal barriers. Services are locked down too tightly. Permissions become political. As a result, the people closest to the workflow often cannot combine the information they need to improve it.

That is a useful challenge to the usual enterprise AI script. Most leaders talk about models, copilots, or governance first. Joshi points instead to access and enablement. In his view, companies get better returns when they unlock the right internal data for more people, not just for technical teams.

“We basically unlocked data for everyone first.”

He used his day-job experience at DeepL as the example. There, he said, internal teams across design, marketing, and sales can use AI models and combine data from systems like Salesforce and user interview tools to build stronger customer profiles and improve sales execution. In practice, that means sales, design, and marketing teams can combine customer signals without waiting for engineering to build every workflow for them. The broader point is what matters for UC Today: AI starts to look more valuable when it supports cross-functional decision-making instead of staying trapped in technical silos.

Trust Matters More Than Tools

Joshi also linked AI success to organisational culture. If teams are constantly fighting over permissions, ownership, or access to code and systems, he said AI will magnify those tensions rather than solve them. In that environment, agents and copilots do not create speed. They create more organisational conflict.

“If you don’t have trust, then you’ll have a lot of organisational conflict.”

That is a sharp point for UC and productivity buyers. Many AI projects still fail for non-technical reasons. Teams do not trust each other enough to share context. Leaders do not trust employees enough to open access. Staff do not trust the system enough to depend on it. That is why governance has to be more precise than simply locking everything down.

Joshi drew an important line here. Companies should not expose sensitive production data carelessly. But they should not confuse sensible control with blanket restriction either. Revenue data, deal signals, sales intelligence, and workflow information can still help non-technical teams make better product and business decisions when handled properly. That aligns with a wider shift toward more structured enterprise AI access models, whether through enterprise AI platforms or internal governance layers built around approved data sources.

The Real Investment Priority Is People

When asked where CIOs should invest over the next 12 to 18 months, Joshi did not answer with infrastructure, model labs, or another platform category. He answered with people. Give employees access to tools that genuinely help them do the job. Identify the internal champions already excited about AI. Let them teach others. Build workshops. Build communities. Let the knowledge spread.

That is an important correction to the current market. Too many companies still spend heavily on AI ambition while underinvesting in AI adoption. The risk is not just that AI replaces people. It is that companies overfund the technology and underfund the adoption layer that makes it work. Joshi warned that this imbalance is already creating a more troubling outcome: not a clean story of AI replacing humans, but a messier one where companies spend so much on AI that they leave less room for the people needed to make it useful.

For enterprise buyers, the takeaway is blunt. AI stops being hype when more employees can use it safely, more teams trust one another enough to work with it, and leaders invest in enablement as seriously as they invest in tooling. The companies that win with AI will not be the ones with the best model. They will be the ones where people can access the right data, trust the system, and know how to use it.

FAQs

What did Akash Joshi say companies should do to move beyond AI hype?

He said companies should put AI tools into the hands of non-engineers and unlock the right internal data so more teams can use AI to improve real workflows.

Why does trust matter so much in enterprise AI?

Because without trust between teams, organisations create bottlenecks around permissions, code access, and data access. That slows down adoption and weakens ROI.

How should companies balance AI access and governance?

Joshi’s view is that companies should protect sensitive production data, but still give employees access to useful business data that helps them make better product and revenue decisions.

What should CIOs prioritise over the next 12 to 18 months?

He argued that leaders should invest in people, internal champions, workshops, and AI communities, not just in tools or model development.

Why do some companies benefit from AI more than others?

Because they combine practical access, better internal trust, and stronger enablement. They focus on making AI useful across the organisation rather than restricting it to a few specialists.