ChatGPT is helpful, but it is not all knowing. Cash for Keys needs proper legal advice.

Recently, I read an article about how difficult it has become for attorneys when clients or pro pers (those who represent themselves without an attorney) start to get answers from AI, such as ChatGPT, that are wrong. This is not entirely surprising to me, but it now has finally made its way to my day to day.

When I negotiate tenant buyouts I mainly do so directly with tenants. There are of course deals that are easy, deals that take a lot of back and forth, and deals where we are talking in circles. Now I have deals where I am not discussing reality with a tenant, but am fighting AI generated responses.

With AI, you can put in what you want it to know so that it gives you what you want to hear, and in your own voice or style of speaking. That is great if you want to live in an echo chamber, but not great when you are having a real life discussion that involves varies laws. The law has nuance. A local law layers over a state law that may parallel a federal law. You can have a local law that references another local law. You can have one local law that supersedes another local law. You can have a law that is now outdated. You can have a law that applies here, but does not apply there. When you tell AI what you think applies to you, or that you want to classify a person or situation one way, it will respond, but you can be setting it up to fail. In turn, you are getting unreliable information that sounds great for you, but is not reflective of your reality.

Where I struggle is trying to explain to someone what the internet told them is not right. This happens a lot when one jurisdiction has a higher relocation fee than another jurisdiction. A tenant clearly lives in one, but wants to argue they are entitled to the amount outlined by a completely different city. Or I have one person alleging liability based on a case AI told them about, which is entirely unrelated to the set of circumstances at hand.

I've been called all sorts of names, including incompetent, because what I said did not fall in agreement with what someone else wanted. It is not incompetence, but it does pose a lot of challenges for legal professionals. We now have to combat AI being a cheerleader for the user. How can I manage expectations when AI is undoing them? AI, that has no feeling, just directive.

As of late, I believe ChatGPT will no longer be giving legal or medical advice, and rightfully so. While I have used ChatGPT to help me organize thoughts and even plan travel, there is not presently a place for it to be "practicing" law. There have been a few attorneys sanctioned for relying on AI and not fact checking the cases, which turned out to be fake.

In the end, I cannot tell anyone how to do, but I caution everyone, when you do research you have to go to the source, which takes time and a lot of dry reading. Shortcuts are just that, shortcuts. They can help put you in the right direction, but AI at this point is not determinative and should be used lightly.

The information in this post is for general information purposes only. Nothing on this post should be taken as legal advice for any individual case or situation. This information is not intended to create, and receipt or viewing does not constitute, an attorney-client relationship.