How AI Is Writing Better Contracts Than Most Solicitors

The average SME contract was written by a solicitor in 2019 and has been used ever since. It’s a template from a previous engagement, adapted with find-and-replace. The parties’ names have changed. The specific scope hasn’t. The payment terms are generic. There are clauses that don’t apply to this situation and missing clauses that should definitely be there.

This is the reality of contract management for most small businesses. Either you pay a solicitor £300–500/hour to draft something bespoke, or you use a template that was probably designed for a different situation than yours.

Why templates fail

Templates are one-size-fits-all solutions to a problem that isn’t one-size. A contract for a one-day branding workshop is structurally different from a 12-month IT support agreement, even if both are ‘service contracts’. When you use the same template for both, one of them is badly served.

The three most common failures we see in SME contracts:

• Vague scope of work — ‘design and development services’ means nothing. How many revisions? Which deliverables? What does ‘complete’ look like? When scope is ambiguous, disputes are inevitable.
• Missing acceptance criteria — if the contract doesn’t define what ‘satisfactory completion’ means, any supplier can claim they’ve finished and demand payment.
• No IP clarity — who owns the code, the designs, the data? If it’s not written down, it’s assumed to be the supplier’s. Many buyers don’t find this out until the relationship sours.

What AI does differently

Bundle IQ’s contract generation doesn’t start from a template. It starts from the deal. The system reads the procurement request, the winning vendor’s proposal, the agreed price, the milestone structure, and the delivery timeline. Every contract clause is generated in response to those specific facts.

This means the scope of work section actually describes the work. The payment terms match the agreed milestones. The IP clause reflects who is creating what. The termination conditions are calibrated to the contract duration. The liability cap is set relative to the contract value.

And because the model has been trained on thousands of commercial contracts across dozens of categories, it knows what clauses are typically contentious in IT support agreements that aren’t relevant in construction contracts, and vice versa.

The plain-English layer

One thing AI can do that a standard solicitor’s output can’t: explain itself in plain English. Bundle IQ generates every contract with a two-paragraph plain-English summary at the top and annotated key terms in a side panel. A non-lawyer can read the summary in two minutes and understand exactly what they’re signing.

This matters enormously for SMEs, where the person signing a contract is often the same person running the business — not a lawyer, not a procurement specialist, just someone who needs to know: what am I committing to, what are my protections, and what happens if things go wrong?

What AI can’t do (yet)

AI-generated contracts aren’t right for every situation. For high-stakes or genuinely novel agreements — a joint venture, an IP licensing deal, a major acquisition — you still want a specialist lawyer involved. The nuance and judgement required in those situations goes beyond what current models can reliably provide.

For the commercial service agreements that make up 95% of SME procurement — IT support, facilities, marketing, logistics, trades — AI-generated contracts are not just ‘good enough’. They are, in our experience, frequently better than what businesses were using before.