Pitfalls, Perils, and Playfulness: Exploring the Fun and Risks of AI in the Legal World*
Artificial intelligence (AI) has become increasingly prevalent in our daily lives, from chatbots like ChatGPT to virtual assistants like Siri and Alexa. While AI has the potential to revolutionize industries and improve efficiency, it also comes with its own set of perils. In Pennsylvania, there are legal concerns surrounding the use of AI that should be taken into account by individuals and businesses alike.
One of the primary legal concerns with AI is privacy. AI systems are often designed to collect and analyze large amounts of data, which can include sensitive personal information. Under Pennsylvania law, individuals have a right to privacy that extends to their personal information. This means that businesses that use AI systems must ensure that they are collecting and using personal information in a way that is consistent with state and federal privacy laws…
STOP!! Stop reading. I didn’t write any of that. But it sounds like something I would write, doesn’t it? That’s because I asked ChatGPT to “write a legal blog using Pennsylvania law about the perils of AI like ChatGPT. The tone, syntax, and style should be similar to the blogs found here and here.” (Those links are to blogs that I have previously written).
While there are countless issues with AI, including privacy, as pointed out by ChatGPT, one of the biggest problems is that for most of us, it is almost impossible to tell the difference between content generated by AI and content generated by a human. Bad actors have already started to use AI to take advantage of this and to scheme victims into disclosing sensitive information.
For example, one emerging pattern is the use of voice-generated AI to convince unknowing people that a family member is in danger and needs money. In that situation, criminals are using AI software to call someone pretending to be a family member, using what sounds like the actual voice of the person they are purporting to be. When the caller states that they are in trouble and need money wired right away, the victims are falling into the trap and sending the money. By the time they realize their family member is okay and that it was a trick, the money is long gone.
Other problems arise with AI due to its limitations. While it seemed like a simple solution to ask ChatGPT to write a blog to save me some time, the information that it provided was not completely accurate. While ChatGPT was able to recite some law, it was not able to fully comprehend and apply Pennsylvania-specific law to the set of facts that I entered. Hence, the problem. For a non-lawyer looking for legal advice, it is impossible to be sure that the advice generated by AI is accurate. This can lead to costly problems down the road for anyone who follows incorrect legal advice.
I must admit that playing around in ChatGPT is a lot of fun and I am getting a kick out of coming up with obscure requests, which, for the most part, have been satisfactorily met. Asking ChatGPT to generate a packing list for a week-long trip to the beach with 3 kids under 6 has definitely made my summer vacation planning easier!
My advice for anyone looking to explore what modern AI has to offer is to be cautious about the information you are receiving. And, as always, to always second guess anything that seems too good to be true, too convenient, or just doesn’t feel right.
* I didn’t write this title either. I asked ChatGPT to “write a title for a legal blog about AI using the words pitfalls peril and another word that starts with P that means fun”.