Why Your Leaders Still Think AI Is Just a Support Tool

I see this all the time with UK SMEs. The board hears "AI" and thinks chatbot. Something that answers questions, drafts an email, maybe summarises a meeting. A clever assistant sitting in the corner. Harmless. Helpful, even.

Then someone on the shop floor or in finance quietly builds an agent that actually does things — chases invoices, reconciles stock, flags dodgy purchase orders — and leadership gets a bit nervous because nobody planned for that.

That gap between "AI as a support tool" and "AI as something that acts" is exactly what McKinsey found in their State of Organizations 2026 report, and it's worth understanding before it catches your business out.

What the research actually says

McKinsey surveyed over 10,000 leaders between June and September 2025 for their State of Organizations 2026 report. Two findings jumped out at me:

• Most respondents expect AI to act mainly as a support tool in the near term. Something that helps humans work, not something that works on its own.
• Younger leaders are significantly more optimistic about autonomous AI roles. The people who'll be running your business in ten years think agents will be doing proper work, not just assisting.

So you've got a generation of senior leaders quietly assuming AI stays a helper, while the next generation down is assuming the opposite. Both groups are making plans based on completely different versions of the future.

Why this matters for your business

Here's the thing — this isn't just an interesting survey nugget. It shows up as real decisions inside SMEs:

• You buy a chatbot licence when what you needed was an agent that chases late invoices automatically.
• You write an AI policy that bans autonomous actions, then wonder why nothing gets faster.
• Your younger managers get frustrated and start building workarounds with free tools nobody's governs. That's your data walking out the door.

The support-tool mindset isn't wrong, mind. It's just incomplete. AI is a brilliant support tool today. But the trajectory is clear, and the firms planning for both — assistant now, agent later — are the ones who won't get caught out.

What I'd actually do about it

Three practical steps, whether you're a 20-person manufacturer or a 200-user group:

• Use AI as a support tool now, in one painful process. Pick something measurable — quote turnaround, invoice chasing, report drafting. Get a real win before you get ambitious.
• Fix the data before you let anything act autonomously. An agent with bad master data doesn't just give wrong answers, it takes wrong actions. Clean part numbers, customer records and definitions first.
• Ask your younger managers what they think AI should do. Seriously. McKinsey's data says they see the future differently — and they're usually the ones already experimenting. Better to channel that than ban it.

The mistake isn't being cautious about AI. The mistake is assuming today's version is the final version. It isn't. Plan for a helper now and a teammate later, and you won't have to rip anything up when the change comes.

References & Further Reading

1. McKinsey, The State of Organizations 2026 (survey of 10,018 respondents, June–September 2025)
2. MIT Sloan Management Review & BCG, The Emerging Agentic Enterprise
3. Anthropic, Building Effective AI Agents

Planning your AI journey?

I help UK SMEs get AI-ready the sensible way — clean data first, one workflow at a time, then agents that actually earn their keep.

Let's have a chat

Matty Hatton is the founder of Digital Adaption, an ERP and data consultancy based on the Wirral. He has spent 15 years delivering ERP transformations for manufacturers, including leading the data migration on a £4.5m consolidation of four legacy systems onto a single Infor LN cloud instance for a 220-user group. He holds an MSc in Digital Transformation and IT Strategy from Manchester Metropolitan University and is Microsoft PL-200 certified.

LinkedIn | Get in touch

Start with a 30-minute data risk call