Craddock v. OpenAI: Suit Over ChatGPT's Faulty Contract Advice

CategoryGovernanceAccuracyReportedSource ↗

Status: allegations only. The case is Craddock v. OpenAI OpCo LLC, 3:26-cv-06920 (N.D. Cal.), reported by the Daily Journal on July 9, 2026. We did not find OpenAI’s response, a motion to dismiss, or any ruling for this case. Check the court docket for current status.

According to the report, the plaintiff bought a premium ChatGPT subscription in 2025 for business strategy and document drafting. The complaint alleges OpenAI presented ChatGPT as a capable business assistant that gives reliable advice, and that she relied on defective confidentiality agreements and other inaccurate guidance. The complaint argues a paying customer could reasonably expect accurate, appropriately qualified answers on legal and contractual questions.

A separate suit, Nippon Life v. OpenAI, makes related claims about ChatGPT and legal advice, and OpenAI has asked a federal court to dismiss a legal-advice suit on the ground that ChatGPT is not a lawyer.

Why it matters for agent builders: confident, fluent output in a regulated domain invites reliance. The question courts are being asked is whether disclaimers are enough when the product is marketed as a capable assistant.

The fix: state the limits of the product where the advice is given, route legal and medical questions to qualified review, and make sure marketing claims match what the system can reliably do.

Failure Patterns Behind This Incident

Cite this incident

Agent KB Team. "Craddock v. OpenAI: Suit Over ChatGPT's Faulty Contract Advice." AI Agent Failure Patterns, July 9, 2026, https://agent-kb-autopublish.vercel.app/incidents/craddock-v-openai-chatgpt-business-legal-advice/. Accessed October 15, 2026.