ChatGPT Shouldn’t Be Responsible For Your Estate Choices

It doesn't matter if you've called your ChatGPT account Chad, Chip, or Claude (or some other weird anthropomorphism), it's not a real person.

It doesn’t know you.
It doesn’t understand your family.
It cannot judge nuance.
And it absolutely cannot give legal advice tailored to your personal circumstances.

It produces words based on prompts, which is very different from producing a legally effective estate plan or legal advice.

Being informed is wonderful. I fully support curiosity, questions, and a healthy interest in understanding how things work. What’s risky is when online, machine based research quietly morphs into documents and decision-making without advice.

Truth bomb: convenience and correctness are not the same thing. And when it comes to wills, the difference matters enormously.

The Cost Of ChatGPT's "Free" Advice

Lawyers (including me) are increasingly being asked to defend their advice against something a client has generated or reviewed through AI, as though “free advice” from a machine should be treated as equal or better than years of experience, study, and a deep understanding of both the law and the circumstances of the client sitting in front of them.

Estate planning is not just about knowing the law. It’s about:

• Seeing patterns from hundreds of past estates.
• Understanding how families actually behave (not how we hope they behave).
• Anticipating disputes before they arise
• Recognising emotional, financial, and psychological dynamics that never appear in a template.
• Applying judgment to nuance, which is the fundamental thing technology struggles to interpret.

A machine can generate information. It cannot replace professional judgment. And it certainly cannot sit across the table from a client, listen to what’s not being said, and guide them through decisions that affect generations.

Just like I've read a lot of true crime novels and watched a lot of Grey's Anatomy,  but that doesn't make me qualified to be a Detective or a Surgeon.

What ChatGPT Can't Do

I've discovered AI can help me write letters that are devoid of phrases people, apparently, find abrasive. It helps me search for phrases, find articles, and bullet point complex matters. AI helped me generate a list of possible birthday party ideas for my 13-year-old. ChatGPT could interpret a friend's recent blood test and write a menu plan that takes everyone's likes, dislikes, and food disabilities into consideration.

I'm not such a Luddite that I don't see the benefits, and recognise this is the way of the future. But at what cost?

RISK 1: AI Can’t Assess Capacity or Undue Influence

To make a valid Will, you must have testamentary capacity. You need to understand:

• What assets you have

• Who might reasonably expect to benefit

• The consequences of the decisions you are making

An AI tool cannot assess whether you meet that threshold. It can't tell if someone is pressuring you or notice hesitation. It certainly can't say, “Is this really what you want?”

A lawyer’s role is not just to write things down. It is to ensure the legal decisions you make are truly yours, made freely, knowingly, and properly.

RISK 2: AI Doesn't Tell You What You've Missed.

When we meet, I'll ask questions you didn’t even know were important:

• Are there stepchildren?

• Is this a blended family?

• Do you have an estranged relationship?

• Are you legally divorced or just separated?

• Do you have a beneficiary with a disability?

• Are any of your beneficiaries vulnerable, or perhaps just financially irresponsible?

• Is anyone likely to challenge the will?

These aren't just questions I ask because I'm nosy. They shape how a will must be structured. AI only works with what you put into it. It can't identify risks hiding in what you didn’t say.

In my experience, the biggest problems people encounter are usually the ones they didn’t know existed. How do you ask ChatGPT to factor in problems you don't even know have the possiblity of being a problem? (hint: you don't)

RISK 3: Online Legal Documents Don't Manage Dispute Risk

A properly advised estate plan is not just about distribution. It is about risk management. AI gives you information, not strategy. When we meet, I will consider:

• Who might bring a claim against your estate.

• How to structure gifts to reduce disputes.

• How to clearly document your intentions.

RISK 4: AI Doesn't Consider Complex Assets

Many people don’t just “own things.” They control them through:

• Companies

• Family trusts

• Business structures

• Superannuation arrangements

These assets don’t always fall into your estate automatically. If they’re not dealt with correctly, your will may not control them at all. A document that says “leave everything to…” is meaningless if the asset never becomes part of the estate in the first place. AI doesn't know how your assets are owned (and it doesn't know to ask).

RISK 5: AI Reduces Estate Planning To A "Simple" Task, When It Rarely Is

My inbox is full of emails that start like this.... "Fiona, can you just do this simple thing that AI suggested...?"

Even the most straightforward estate is rarely straightforward. AI has a habit of reducing things to their simplest form without taking all the layers and what-ifs into consideration.

Estate planning is not about filling in blanks. It’s about avoiding avoidable problems by asking the right questions, the important questions, BEFORE there’s a problem.

Good legal advice doesn't rely on YOU knowing the "right" questions to ask. You get asked the questions and then a strategy is created in response to your answers.

Knowledge At What Cost?

I totally get it, we don't know what we don't know and AI seems like such an easy, accessible, and inexpensive way to gain knowledge. But there are costs.

Estate plans have their moment to shine after you're no longer here to explain what you were trying to achieve. If you leave a will that provides more questions than answers you're not leaving a gift for your family, you're leaving something lawyers and familiies will fight over.

Uncertainty = conflict.

Making sure there's no uncertainty is less about law and more about finding someone you can trust to help you look after the people you love.

As a real-life, human lawyer, I can do that better than any Chad, Chip, or Claude (or whatever you want to call your AI program of choice).