The option to quickly create a contract for your business needs on AI platforms such as ChatGPT or Claude is quick, cheap and immediately accessible. AI follows patterns, uses information that has been shared before and generalises in a very slick format that offers immediate reassurance. But what can go wrong? Well, a lot. In my role as a commercial lawyer, I am increasingly having to unravel makeshift AI business contracts that are exposing my clients to hefty legal fees.
Generative AI tools offer fast, polished drafts, but using them without professional oversight introduces severe commercial risks.
We understand that lawyers fees are expensive for founders starting out, so myself and my colleagues created shizl – affordable and accessible tailored legal documentation platform.
Here are three examples of what can happen if an AI creates crucial business contracts and gets them wrong.
- Employment Contracts: Employment law relies on local law, statutory update cycles, and specific mandatory terms such as precise notice periods, statutory rights, and working hour restrictions. The expense that a generic AI contract can cause lies in it mixing up UK and US employment law; this means the contract is unenforceable and will fail to properly outline what the responsibilities of the employer and employee are. Without stated responsibilities, employees can bring constructive dismissal claims, face an employment tribunal and huge mandatory compensation payouts.
- IP Documents: Intellectual property law requires precise language to successfully transfer ownership from a developer, contractor, or employee to your business ownership. AI models frequently draft generic “assignment” clauses that fail to legally convey future IP or rights under specific regional statutes. If you attempt an exit, seek venture funding, or try to sell your company, an audit will flag flawed IP documentation. I have witnessed this firsthand, and it is devastating for all concerned. Revising ownership gaps under duress or losing an acquisition deal entirely costs tens of thousands in legal fees.
- International Data Transfer Contracts: Transferring personal data across borders requires strict adherence to valid mechanisms, such as Standard Contractual Clauses (SCCs) or the UK International Data Transfer Agreement (IDTA), alongside mandatory Transfer Impact Assessments (TIAs). AI models generate plausible-sounding “data clauses” that lack valid statutory mechanisms. Operating with non-compliant data transfer agreements exposes your business to regulatory enforcement from authorities like the ICO or EU DPAs, where fines can reach up to 4% of annual global turnover or €20 million.
Why Legal Platforms and Lawyers Cost Less Overall
While hiring a specialist lawyer or using established legal template platforms requires an initial outlay, they offer critical safeguards:
- Lawyer Vetted: Purpose-built platforms such as shizl provide templates drafted and maintained by practising lawyers, ensuring compliance with current statutory frameworks.
- Liability & Insurance: Solicitors carry professional indemnity insurance. If a drafted document causes a loss due to a legal oversight, you have legal recourse; with an AI model, you bear 100% of the liability.
- Enforceability: Proper legal documents are engineered to hold up in court, preventing protracted litigation costs before disputes arise.
Saving money on AI legal drafts frequently leads to five- or six-figure corrective costs tomorrow. For core operational documents, relying on legal professionals or specialised legal platforms remains the most cost-effective decision for long-term protection.
