The dangers of AI drafted contracts: what GenAI gets wrong in commercial agreements

From the perspective of a small business owner, the potential benefits of using generative artificial intelligence to draft legal documentation seem compelling at first blush. There are a growing number of AI platforms purporting to offer comprehensive off-the-shelf legal documentation at a mere fraction of what traditional law firms charge – all at the click of a mouse. Indeed, these AI tools provide almost instant output and, on face value, produce legal documents that looks impressive and fit for purpose, with plenty of ‘legal’ phrases and jargon.

What’s the catch? Well, as the saying goes, what looks too good to be true often is, and in the case of using AI to draft commercial contracts and corporate documentation, the potential pitfalls far outweigh the purported benefits. Indeed, although to the untrained eye AI drafted legal documents appear to be more than adequate for any SME, they inevitably fail to cater for the specific needs and wants of business owners and neglect the nuances of the particular industry they operate in. In more unfortunate cases, they produce documents that omit vital provisions that leave business owners carrying unknown risk and jeopardy. What might seem good value at the outset could lead to severe financial consequences in the long term.

Although this blog is relevant to AI-generated commercial contracts and corporate documentation more broadly, we have focused below on some specific problems that our litigation teams have encountered in regard to shareholder agreements and share purchase agreements. These are two critically important and common documents that, when poorly drafted, present business owners with a range of potentially severe consequences.

Share Purchase Agreements (SPAs)

A share purchase agreement (SPA) is a legally binding contract that sets out the terms and conditions under which a buyer agrees to purchase shares in a company. It is a key document in private company sale transactions and is typically used in mergers and acquisitions (M&A) involving the transfer of ownership of a target company.

SPAs cover various critical elements of the transaction, including the purchase price, payment mechanisms, conditions of sale, and arrangements for completion. In short, it is well worth shareholders investing the time and money in ensuring that the SPA is properly drafted. If it isn’t, then expensive and time consuming litigation may follow.

In the relatively short period of time AI legal tools have been used as a drafting aid to produce SPAs, we have encountered the below key pitfalls:

• Warranties being poorly drafted (too widely by sellers and too narrowly by buyers), leading to sellers accidentally warranting something is true when they can’t possibly know, or sellers not having the full benefit of warranties which would have been standard;
• Warranties purporting to cover factors that cannot legally or factually actually be warranted (these clauses are redundant at best and could result in costly litigation at worst);
• Payment terms being so unclear that they risk being unenforceable (this is particularly detrimental in relation to deferred consideration, so delayed payment or earn-out provisions);
• References to foreign/outdated law (this risks a party contracting out of the jurisdiction of England and Wales, and into some foreign, complex, and expensive jurisdiction).

Shareholders Agreements (SHAs)

A shareholders’ agreement is a private, legally binding contract between the shareholders of a company and potentially the company itself. It governs the relationship between the shareholders, their rights and obligations, and the management and operation of the company.

The agreement typically addresses key matters such as voting rights, the issuance/transfer of shares, the appointment of directors, and the distribution of dividends. It may also include provisions that limit the powers of the board of directors and designate certain issues as reserved matters (requiring shareholder approval).

We have recently seen clients involved in the following mistakes when relying on AI tools draft Shareholders Agreements:

• Not dealing properly with share class rights and/or confusing/conflating share classes (this could unintentionally significantly alter the value and utility of certain shares issued, potentially rendering a person’s shares useless or worthless);
• Drafting bad leaver provisions so poorly that an underperformer can’t be forcibly removed from the company (this can result in very costly litigation and/or significant payouts); and
• The agreement failing to account for the key risks in a given industry/business relationship (this could effectively remove any purported benefit of drafting the agreement in the first place).

In short, it is hard to imagine AI ever being able to replace the expertise and oversight of a solicitor and the advantages of using AI to draft key agreements is far outweighed by the potential pitfalls. AI is particularly poor at exercising legal judgment and interpreting legal risks, which are important from a commercial standpoint.

The key risk factor remains in the SME world – as more and more SMEs, in a tight economy, turn to ChatGPT to ‘save money’ drafting contracts. In our experience, however, the saving of a few hundred pounds in not instructing a solicitor to draft your agreement, will often result in the spending of a few thousand (possibly a few hundred thousand, depending on the dispute and size of the company) pounds on litigation to resolve the errors which could have been prevented at the outset.

How can Ellis Jones help?

As a full service law firm, we are able to assist your company in the drafting of both corporate documents and commercial agreements, as well as provide advice and representation in the event that problems arise.

Our specialist team of Commercial Dispute Resolution Solicitors have extensive experience in interpreting the risks in corporate and commercial documents. To contact a member of the team call 01202 525333 or fill out our enquiry form below.