The Operator’s AI Roundup — June 2026

The AI Tools Business Owners Are Actually Using Right Now — and What Changed This Month

Every month I publish what I’m seeing on the ground. What’s working in our own firm. What owners I’m talking to are actually using. What changed. What didn’t matter as much as the headlines said. This is the June 2026 edition.

I run an M&A advisory firm in Nashville. We also build AI operating systems for business owners. That means I’m living inside this technology every day, using it on real client work, real deals, real business problems. Not benchmarking it. Using it.

The tools are moving fast. Faster than most business owners have time to track. So every month I publish what’s actually worth knowing — what changed, what we’re using, what I’d recommend to an owner who wants a practical view without the hype.

What We’re Using Inside Icon Right Now

Claude (Anthropic) — Our Core Intelligence Engine

Claude is what runs our firm’s operating layer. Every deal brief, every client research package, every proposal, every follow-up sequence — Claude is in the middle of it. We’re not using it as a writing assistant. We’re running it as an operating system. There’s a meaningful difference.

What changed in June: the agentic capabilities have gotten significantly better. Claude can now coordinate multi-step workflows with less supervision, which means the orchestration layer we build for clients is more capable than it was 90 days ago. If you haven’t revisited what Claude can do recently, it’s worth another look. It’s not the same product it was six months ago.

Notion — The Intelligence Layer’s Home Base

Every piece of institutional knowledge in our firm lives in Notion. Deal intel, client histories, process documentation, decision frameworks. It’s not glamorous, but it’s load-bearing. AI is only as useful as the information it has access to. Notion is where we make our knowledge accessible.

What changed: Notion’s AI layer has improved significantly and the MCP integration now makes it accessible programmatically. For businesses building a real intelligence layer, the combination of structured Notion knowledge and AI that can query it is genuinely powerful. This is the architecture we install for clients.

HubSpot — The Orchestration Layer’s Execution Surface

CRM done right is an orchestration tool, not a contact database. We use HubSpot as the execution surface for our client and prospect orchestration layer. Sequences, follow-up automation, deal pipeline intelligence — when AI can read from and write to your CRM, your business development layer starts running without constant owner input.

If you’re a business owner with HubSpot and you’re using it as a contact list, you’re leaving the most valuable part of the tool on the table.

What Changed This Month That Business Owners Should Know

Agentic AI Is Becoming Real

The phrase “AI agents” has been overused to the point of meaninglessness. But something real is happening underneath the hype. AI systems are increasingly capable of running multi-step processes with minimal human supervision. Research that used to take a half-day takes 20 minutes. Follow-up sequences that had to be manually triggered now run on conditions. Analysis that required a dedicated person to assemble is assembled automatically.

For business owners, this means the orchestration layer — the systems that run your business without you remembering to run them — is becoming dramatically more capable. What was difficult to build 18 months ago is now accessible to a $5M business with the right implementation partner.

The Gap Between Owners Who’ve Started and Those Who Haven’t Is Widening

I talk to business owners every week. The ones who started building their intelligence infrastructure 12 to 18 months ago are operating differently now. Their capacity per person is higher. Their owner dependency is lower. Their decisions are more consistent. The gap between them and comparable businesses that haven’t started is real and growing.

This matters for valuation. A buyer looking at two similar businesses in the same market will price the more intelligent one higher. Not because AI is a premium feature. Because owner dependency is a discount factor, and the business that’s eliminated it is simply less risky to buy.

The Tools Are Cheap. The Implementation Is the Work.

The most common thing I hear from business owners is that they don’t know where to start. The tools are overwhelming. There are hundreds of them. Every week something new launches. Every vendor claims their product is the answer.

Here’s the honest truth: the tools are almost irrelevant. Claude, ChatGPT, Notion, HubSpot — these are all capable of supporting a serious intelligence layer. The work is in knowing what to build, in what order, and how to make it actually run in your specific business. That’s not a software question. It’s an operating question. And it’s the question we help owners answer.

What I’d Tell a Business Owner Who’s Ready to Start

Don’t start with tools. Start with an honest look at where your business’s intelligence lives right now. What would stop working if you weren’t there? What decisions require your judgment that shouldn’t? What knowledge exists only because the right people happen to still be with you?

That’s the audit that tells you where to build first. It’s also what we do in the AI Assessment — map every intelligence gap in your business, rank them by impact, and give you a clear picture of what it costs to leave them closed and what it’s worth to open them.

The owners I work with who moved earliest are in the best position now. The ones who waited are still getting there, but the gap is real. If you’ve been watching this space and thinking it’s almost time, the honest answer is that time was 18 months ago. The second-best time is now.

The next Operator’s AI Roundup publishes in July. If you want it in your inbox, reach out and let us know. — Daniel Askew, Founder, Icon Business Advisors