Why business owners should be wary of AI-drafted shareholders’ agreements

A growing number of small businesses are turning to AI tools to draft legal documents, and lawyers are warning it is a false economy that could cost far more than it saves.

Barcan+Kirby’s corporate and commercial solicitors have seen a steady uptick in ‘Google law’ over the past few years, with a notable rise in recent months. This is business owners using AI to draft complex legal documents, including shareholder agreements, before asking a solicitor to give them a quick one-over, or not at all.

While there is no question that AI is useful, there are several risks associated with using it to draft legal documents, as the team discovered in one recent case. A client came to the firm’s Corporate and Commercial team with an AI-generated shareholders’ agreement for a company they had just invested in. On the surface, it looked impressive and used legal jargon. They just wanted a lawyer to double-check it.

However, after a closer inspection, it contained inconsistent definitions, missing cross-references, conflicting clauses, outdated legal concepts, duplication and commercially unworkable provisions. For a start, the shareholding percentages added up to more than 100 per cent, an error basic enough that even a non-lawyer might have spotted it. More significantly, the agreement failed to properly address how directors and shareholders would interact in company decision-making, a gap that could have caused real disputes further down the line. The team included the necessary language to enable the terms to be easily enforced and provided the parties with clarity on the details of what they were agreeing to.

"It was like a car that looked fine from the outside, but when you looked under the bonnet, it was missing the engine and a tank of fuel," said one of the firm's lawyers.

But this is not an isolated case. As artificial intelligence becomes more accessible, many people are turning to AI tools to produce first drafts of legal documents, including shareholders’ agreements and heads of terms. At first glance, this can seem cost-effective, but this approach can create more work and cost than expected.

Because AI models are trained on wide-ranging data rather than tailored to any jurisdiction, they can blend concepts from multiple legal systems or include clauses that do not reflect current law, regulatory requirements or market practice. AI tools respond directly to the prompts they are given, so if instructions are vague, incomplete or internally inconsistent, the output will be too. Lawyers must then pick apart the document line by line, correcting errors, filling gaps, and restructuring the agreement so that it works in practice, which can take significantly more time than drafting a tailored agreement from scratch, particularly where the AI has confidently included provisions that are subtly but critically wrong.

Crucially, using AI to draft an agreement will also mean that, if an issue arises down the track which was due to a fundamental error in the drafting, there will be no recourse for the client. A law firm carries responsibility for the advice it gives and the documents it produces. AI does not.

There are other risks too. AI can present legally incorrect or contradictory terms in a persuasive tone, producing a document that reads smoothly but is internally inconsistent. Errors that are not obvious to a non-lawyer can leave a business exposed to real legal or commercial risk, especially if the agreement later needs to be enforced. Reviewing and correcting an AI draft can also end up costing more than producing a bespoke agreement would have in the first place, as the client in this case discovered. The firm's team has also seen AI confidently produce its own incorrect answers to legal queries, only admitting the error once a qualified person prompts it to reconsider with more detailed analysis.

Lawyers are not arguing businesses should avoid AI altogether. Used well, it can help owners think through the issues they want covered, generate an early bullet-point structure, or organise their thoughts before speaking to a solicitor. For clients who choose to use AI as a starting point, one helpful step is to share the original prompts with their lawyers. These prompts often reveal the client’s priorities, concerns and assumptions, information that may not be obvious from the draft alone. Seeing the prompts can help lawyers quickly identify key issues and ensure those commercial points are properly captured in the final agreement.

Looking ahead, better integration with verified legal databases, tools that flag uncertainty, and more structured workflows combining AI with lawyer oversight should all help reduce these errors over time. But even as the technology improves, it is unlikely to replace the need for specialist legal judgement, bespoke drafting, or a real understanding of the commercial objectives behind each deal.

The warning comes as more small and medium-sized businesses look to cut legal costs wherever they can. The firm's message to owners considering the same shortcut is straightforward: a shareholders’ agreement is not a document you want to discover is broken after a dispute has already started. Getting it right the first time, with proper legal input, remains far cheaper than fixing it later.