Lawyer Tasha Dickinson receives phone calls at least 3 or 4 times a week from confused clients. They call her for legal advice and find artificial intelligence chatbots like ChatGPT or Claude. Many clients try to hide their use of these digital tools. However, Dickinson easily spots the origin of their questions. She hears their robotic phrasing and notices their sudden interest in strange tax loopholes.
Recently, a wealthy Florida resident with a net worth of over $10 million contacted her office. He wanted to create a community property trust for his family. He told Dickinson that an AI chatbot recommended this specific trust to save his children thousands of dollars in taxes. Dickinson immediately pointed out a massive flaw in his grand plan. The client’s wife died a few months earlier. She reminded him that a community property trust requires a living husband and wife. The client went completely silent on the phone before admitting that the AI thought the trust offered a brilliant strategy. Dickinson told him the strategy only works in a different universe, not in his real life.
Other lawyers who cater to rich families report the same frustrating trend. Robert Strauss works as a partner at a busy law firm. He says his clients regularly upload 50-page trust documents directly into online AI systems. The clients then return to his office with 15 or 20 new questions and suggested edits. This new habit forces Strauss to defend his professional work. He has to sit down and explain why a robot’s generic recommendations actually ruin the client’s specific financial situation.
Strauss notes that addressing these AI suggestions undermines workplace efficiency. He says his team often wastes 2, 3, or even 4 extra hours answering completely pointless questions. If a lawyer charges $500 or $1,000 an hour, this wasted time costs the client a huge amount of money. Strauss states clearly that he has never received a single useful or workable legal suggestion from an AI chatbot. He feels this growing habit slowly ruins the basic trust between a client and their legal team.
Beyond wasted time, lawyers worry deeply about data privacy. When clients paste highly sensitive financial details into language models, they expose themselves to major legal risks. Strauss says his firm currently rewrites its client contracts to include strict new warnings. He wants clients to know that sharing secrets with a chatbot can destroy their attorney-client privilege.
A federal judge made a landmark ruling on this exact issue back in February. The judge decided that a criminal defendant lost his legal protection when he discussed his defense strategy with Claude. Dickinson says this specific ruling keeps her awake at night. She does not fear AI making mistakes. She can easily fix bad math or wrong dates. Instead, she worries over clients waiving their legal privacy just to get a quick answer online.
Dan Griffith, a wealth strategy director at a large bank, echoes these same fears. He warns people never to ask a chatbot how to hide $5 million in assets. He also tells them never to ask AI how to draft a sneaky prenuptial agreement. Prosecutors can drag those chat logs straight into a courtroom and use them against you. Yet, even folks who can afford the best lawyers still love the cheap convenience of AI. Nobody likes paying high fees for professional services, so wealthy clients look for free shortcuts just like everyone else.
Dickinson points out that her clients build highly successful businesses and possess incredible drive. They read constantly and want to understand their money. However, this same drive gives them a false sense of legal knowledge. Years ago, clients brought bad ideas to their lawyers after talking to friends at the local country club. Today, they bring bad ideas straight from their computers.
Ed Renn, another wealth lawyer, shared a story about the dangers of bad prompts. One of his clients demanded a plan to transfer $20 million in assets to his wife without paying any taxes. The client proudly claimed ChatGPT promised him a 100 percent marital tax deduction. However, the client forgot to tell the chatbot one important detail. His wife held citizenship in a different country. The law strictly forbids an unlimited tax deduction for a foreign-born spouse without a very specific, complex trust.
Renn says clients fail to understand that AI follows a simple rule. If you put garbage in, you get garbage out. Chatbots constantly fail to understand tricky international tax laws. They also ignore new updates from the Internal Revenue Service. Griffith adds that machines try too hard to give a simple yes-or-no answer. Real lawyers stop and ask clients deep questions about their families and goals before writing a single legal document.
