OpenAI faces $10.3m lawsuit in Illinois after ChatGPT allegedly provided unlicensed legal advice that led a user to reopen a settled case and file meritless court claims
OpenAI is facing a lawsuit in Illinois alleging that its ChatGPT AI system acted as an unlicensed lawyer, exposing the company to $300,000 in compensatory damages and $10 million in punitive damages. The claim arises from an incident involving Nippon Life Insurance Company of America, where a ChatGPT interaction allegedly provided legal guidance that led a user to reopen a previously settled case and subsequently flood a federal court with meritless filings. The plaintiff argues that OpenAI was practising law illegally in the state of Illinois — a regulated activity requiring licensure under state bar rules — and that the harm caused by the AI's outputs is directly attributable to OpenAI. The case goes to the core of one of the most contested questions in AI and law: whether a general-purpose large language model (LLM) — a type of AI system trained on vast text datasets to generate human-like text — can be held liable for practising law without a licence when its outputs are used in legal contexts. US state bar rules and, analogously, the Solicitors Regulation Authority (SRA) rules in England and Wales prohibit the provision of reserved legal activities by unregulated entities. The case is an early but significant test of how courts will apportion liability when AI systems cross from information provision into what is effectively legal advice.
Why this matters
This litigation is directly relevant to the 'AI and legal practice' debate that is reshaping how Magic Circle and US firms in London are deploying and governing AI tools. If a court holds that an LLM can 'practise law' for liability purposes, it will fundamentally change how firms, clients, and AI vendors structure their terms of use, disclaimers, and liability frameworks. The 'why now' driver is the rapid adoption of AI tools by non-lawyers for tasks that straddle legal and non-legal advice — a pattern that the SRA has flagged as a consumer protection risk in the UK context. For commercial lawyers, the immediate implication is that AI governance policies, vendor due diligence questionnaires, and technology licence reviews need to explicitly address the 'unlicensed practice' risk, particularly for clients deploying LLMs in compliance, contract review, or dispute resolution workflows.
On the Ground
A trainee in a technology or AI governance team would be assisting with vendor due diligence questionnaires for AI tool procurement, drafting AI governance policy sections addressing the unauthorised practice of law risk, and reviewing data processing agreement markup to ensure that AI outputs are framed as information rather than legal advice. They would also assist with regulatory impact assessment memos analysing how the Illinois case maps onto the SRA's reserved legal activities framework.
Interview prep
Soundbite
If ChatGPT can be sued for practising law without a licence, every firm deploying LLMs in client-facing workflows needs to revisit its liability framework today.
Question you might get
“How would you advise a UK law firm on structuring its engagement terms and internal governance to manage the risk that an AI tool it uses in client work could be characterised as providing unauthorised legal advice?”
Full answer
OpenAI faces a $10.3m lawsuit in Illinois alleging that ChatGPT provided unlicensed legal advice that caused a user to file meritless court claims, directly raising the question of whether a general-purpose AI system can practise law. The commercial significance for law firms is acute: if courts accept this framing, AI vendors will face pressure to contractually exclude liability for legal-adjacent outputs, and firms deploying those tools in client work will need watertight terms clarifying that outputs are not legal advice. This fits the broader regulatory trend of authorities — including the SRA in England and Wales — warning about AI-generated legal misinformation reaching consumers without professional oversight. This case will accelerate the development of AI-specific engagement letter clauses and governance frameworks at every firm that touches AI-assisted legal work.
Sources
My notes
saved