The AI bottleneck isn't the tool. It's the gap between the tool and your business.

Last week, Reuters reported that OpenAI and Anthropic have each created joint ventures backed by some of the world's largest private equity firms, with one shared purpose: acquiring engineering services and consulting companies that can help businesses actually implement AI.

OpenAI is raising roughly $4 billion. Anthropic, $1.5 billion. The investors include Blackstone, Goldman Sachs, TPG, Bain Capital and Brookfield. Most of the capital is expected to fund acquisitions of implementation and consulting firms, not model development.

Read that again. The companies building the most powerful AI in the world are spending billions to hire people who can make it work inside real businesses.

That tells you something.

The tension at the heart of enterprise AI

Reuters put it plainly: what is often cast as a high-margin software business "still depends on labor-intensive, highly skilled services."

The reason is straightforward. A powerful AI model trained on the internet does not automatically know how your business works. It does not know your quoting exceptions, your client history, or the three systems your team uses that don't connect to each other. It does not know that your job data lives in one platform, your financials in another, and your scheduling in someone's head.

Making AI do something useful inside a specific business requires someone to understand that business first. To map the workflows as they actually exist. To identify where the data is reliable enough to act on, and where it isn't. To sequence what gets built and in what order.

Blackstone's Jon Gray described this as breaking down "one of the most significant bottlenecks to enterprise AI adoption." He's right. And the bottleneck isn't access to the tools. It's knowing how to apply them.

What this means if you're running a $2M–$25M business

The joint ventures OpenAI and Anthropic are building are designed for large enterprises, PE-backed portfolio companies, and businesses with the complexity and budget to match.

They are not designed for you.

But the problem they're solving is the same one your business has. You have access to Claude, ChatGPT, Copilot, and a stack of workflow automation tools. Your team has probably tried something. A chatbot, an automation, a prompt someone set up and then forgot about.

Most of it hasn't moved the needle. Not because the tools are wrong, but because the implementation was skipped.

The process wasn't mapped. The data wasn't organised. The team wasn't shown how to use the tool in a way that fits their actual day. And when something broke or produced a bad output, no one had a system for fixing it.

The labs are building the deployment infrastructure for enterprise. The mid-market, owner-led businesses doing $2M to $25M, still largely doesn't have that infrastructure. That's the gap.

What good implementation actually looks like

It starts with the operation, not the tool.

Before any AI tool is selected, you need to know which workflows are worth automating, which data is clean enough to trust, and which parts of your team's day are genuinely repeatable. That requires mapping the business first, as it actually runs, not as you think it runs.

From there, tool selection follows the workflow, not the other way around. The right tool for scheduling optimisation in a trades business is not the same as the right tool for proposal generation in a professional services firm.

Then implementation happens in phases: a single workflow, a small team, a short window. You measure what you said you'd measure. You fix what breaks. You expand when it's working.

That's it. It's not a technology project. It's an operations project that uses technology.

The window is open

The labs entering this space is a signal, not a threat. It validates the thesis that implementation expertise is the scarce resource in AI adoption. And for now, the businesses they're targeting are not the businesses most of us are running.

If you've been sitting on AI tools that haven't delivered, or wondering whether your business is actually ready to do something meaningful with this, the answer is probably yes. But you need to start with the operation, not the tool.

That's what we help you figure out.

Book a 15-minute discovery call and we'll give you a straight answer on where your business is and what's worth doing next. Book a Discovery Call →

Source: Reuters, "OpenAI and Anthropic ventures in talks to buy AI services firms," May 5 2026