[NoboruWataya]

Apart from being different jurisdictions, they are different issues. The situation in the article involved a pro se litigant feeding courtroom proceedings to the AI and regurgitating its responses in real time. In that situation you are effectively handing your agency over to the AI. You can't really be said to be representing yourself in any real sense; you are mindlessly parroting what is fed to you.

The situation you describe seems to be more akin to an advanced search or information portal that people can use to guide their self-representation, or even their decision to engage lawyers/discussions with their lawyers (of course, maybe I'm misunderstanding). That stuff has basically always been allowed; nobody is threatening to prosecute Google because pro se litigants use it in their research. There are plenty of websites out there that discuss tenants' rights. There are even template tenancy agreements available online for free.

Also, what were you proposing to use as the knowledge base for your Alexa lawyer? Were you really planning on using ChatGPT or some other general purpose AI? Or would the knowledge base be carefully curated by qualified professionals? And who would create and maintain it, the state? A regulated firm? Or a startup with a name like "DoNotPay"?

[brookst]

Really good thoughts and treatment of the different issues. The line between “tool” and “agent” is blurry and will probably just keep getting blurrier. But I do think it’s important for our judicial system to ensure that any delegation of representation is to a very qualified third party, for both ethical and process/cost reasons.

I’m not sure the startup’s name is especially germane though. If anything, it seems to fit right in with human lawyers like 1-800-BEAT-DUI.