AI Tools 20 July 2026 · Darren Sharples
Solo consultant AI subscription tier suggestion for Anthropic

Here's a problem I've been living with myself, not just training other people through it.

As a solo consultant handling client data, I need a Data Processing Agreement with any AI tool I use. That's not optional. It's the same GDPR obligation I put in front of every firm I train. So I went looking for one on Claude, the tool I actually run my business on, and found there isn't a single-seat option that includes it. To get a DPA, you have to buy a Teams account with two seats minimum.

Fine, I thought. I'll buy the two seats. Except the second seat is a stranger. It has none of my conversation history, none of my project context, none of the working relationship I've built with the tool over months. It's not a spare seat for me. It's a second, empty version of Claude that happens to share a bill.

So I'm paying for compliance and getting a worse tool in return. That's the trade nobody should have to make.

"There is no suitable service within Claude right now that both protects me and enables me to get the best out of the platform."

Why this isn't just my problem

I put this out on LinkedIn expecting a handful of nods. Instead a chartered accountant replied straight away, because her profession carries exactly the same pressure. So did a consultant working in financial services. Anyone regulated, anyone handling client data on their own, anyone who actually reads the compliance requirements rather than ignoring them, hits the same wall.

These aren't people trying to cut corners. They're the ones doing it properly, and the tools are making that harder than it needs to be. Compliance and productivity shouldn't force you to split your own work across two accounts that don't talk to each other.

What actually needs to exist

Not complicated. A single-seat subscription that includes a Data Processing Agreement, with usage limits that match what you'd get on a two-seat Teams account, in one continuous environment. No second seat. No lost context. No paying twice to get half a tool.

I don't know if Anthropic will build it. But if you're a solo consultant or a small regulated practice reading this and nodding, you're not imagining the gap. It's real, and until someone closes it, you're stuck weighing up compliance against actually being able to do your job well. That's not a choice anyone competent should have to make, and it's exactly the kind of gap I help clients spot before it becomes a problem, not after.

Darren Sharples is the founder of Safer Haven AI, an AI risk, training and safety consultancy based in Greater Manchester.

Using AI in a regulated role?

Our AI Risk Assessment maps exactly what you're exposed to, including the gaps in the tools themselves, before a client or a regulator finds them for you.

See how the assessment works