AI is already changing how many businesses work. Used well, it can help with planning, administration, first drafts and internal processes. It can save time, prompt useful questions and allow people to focus more of their attention on judgement, relationships and decision-making.
The risk comes when a useful tool is treated as though it is a professional adviser.
For SMEs, that risk is increasingly visible in documents such as contracts, policies and commercial terms. At first glance, an AI-generated document can look convincing. It may have the right headings, familiar legal language and a structure that appears broadly sensible. The difficulty is that a document can look like a contract without protecting the business in the way the owner assumes it does.
We have seen examples of AI-drafted contracts which appeared sound on the surface but missed fundamental protections. One example involved a limitation of liability clause. The agreement included the heading, and it included wording that sounded relevant, but it did not actually limit liability. Had the client signed it and later faced a claim, they may have found there was no effective cap on their exposure.
That is the kind of issue AI can miss. It may reproduce the shape of a legal document without understanding the commercial risk behind it. A solicitor looking at the same contract will not simply ask whether the wording reads well. They will ask what the client is trying to achieve, where the risk sits, what the consequences may be if something goes wrong, and whether the drafting actually supports the client’s position.
That does not mean AI has no place in business. It can be useful as a planning tool. A business owner may use it to think through the broad structure of a policy, list questions to raise, or prepare for a discussion about what a contract needs to cover. But it should not be used as a substitute for legal advice, particularly where the document will govern money, liability, employment obligations, intellectual property, confidentiality or a long-term commercial relationship.
There is also a practical point around cost. Asking a solicitor to review a badly drafted AI-generated contract can sometimes take longer than asking them to prepare a document properly from the start. Problems have to be identified, unpicked and corrected. In some cases, the drafting may create ambiguities that make the review more difficult than starting with a clear set of instructions.
The risks are not confined to businesses using AI without advice. They are also becoming a serious issue for the legal profession itself.
The courts are increasingly alert to the misuse of AI in litigation. A recent case saw two solicitors and their firm referred to the Solicitors Regulation Authority after a court found that inaccurate legal authorities, apparently generated using artificial intelligence, were included in appeal documents filed before the court. The concern is not limited to regulated lawyers. Litigants in person are also using AI to prepare claims, arguments and references to case law, with judges identifying false citations and inaccurate statements of the law.
This is one of the clearest limits of AI in a legal context. Legal research and case law require accuracy. AI tools can produce confident answers that are wrong, and in some cases can generate fictitious cases or authorities that appear real. Where litigation is involved, that can mislead the court, damage a party’s position and create professional or regulatory consequences for those responsible.
Law firms themselves therefore need clear internal policies on AI use, as well as a culture which recognises both the opportunities and the limits of the technology. Clients are entitled to ask how a firm uses AI, what safeguards are in place, and how legal accuracy and confidentiality are protected.
AI will continue to develop, and businesses should not ignore it. But it should be treated as a tool, not an adviser. Legal advice is not simply the production of words on a page. It is the application of expertise, judgement and accountability to a particular set of facts.
For businesses, the safest approach is to use AI carefully, keep control of what is being produced, and take specialist advice before relying on any document or legal analysis that could affect their rights, obligations or exposure.
Hear more from CJCH Solicitors on the Legal Insights podcast series:
