A long time ago, I studied and practiced law. Once I complained to Renier about the effort of keeping up-to-date with case law. He thought it was funny, because tech changes about a million times faster than law.
The law moves slowly, but even lawyers have laptops. It’s no surprise that there are AI-powered tools for legal research. Many of these are trained on legal documents, like legislation and case law.
400 Bad Request
This week ITWeb reported on the use of AI in a Johannesburg court case. Lawyers had used AI to generate their legal arguments. Their document used citations for non-existent cases to argue its points. After this came to light, the lawyers submitted a revised version. The corrected version contained more fake citations.
There was a similar case in KZN. In both cases, the judge ordered that the Legal Practice Council should investigate the conduct of the lawyers.
This has happened in other countries as well. In 2023, lawyers in the US were fined for using nonsense produced by ChatGPT. In the UK, the court found the legal professionals were negligent. There was also an action for wasted legal costs.
The UK judge stated that legal professionals who use AI have a professional and ethical duty to check the accuracy of research. If they don’t, they may face sanctions, from contempt of court charges to crimilar charges.
426 (Ethics) Upgrade Required
Lawyers and doctors and accountants have an established code of ethics. We expect to be able to trust these professionals. Why? Because if they screw up, they can cause us serious injury or damage.
Our world runs on software. And when software fails, it can cause problems from irritation to death. (Yes, it really can and has.) And yet the software industry has no established code of ethics.
(I know this is a bit of a hobby horse for me. I presented a webinar in 2021 on Programmers, Power and Responsibility.)
599 Catastrophic Failure
(No, there isn’t a 599 server error message.)
Robert Martin is a well-known figure in the software industry. He has championed the need for a professional code of ethics for many years.
Recently I watched a 2024 interview with Martin. He believes that one day a software error will cause a horrible disaster that kills thousands of people. At that point, he says, the world will get very serious about standards.
I hope he is wrong. I hope the tech industry gets its act together before that happens. What do you think? I’d love to hear your comments.