'If you don't use it, then you’re just swimming against the tide… if it's faster and more efficient, people will use it.'
There’s a lot that’s been written about the legal ramifications of using chatbots such as OpenAI’s ChatGPT, but Edwin Morrison, Director at K3 Legal, said none of his clients are rushing to draft policy around its usage.
“I think at the moment, a lot of people are playing with it but once it’s integrated into Microsoft, I think there will be a lot more adoption,” Morrison said, referring to the chatbot’s parent company.
Morrison believes Kiwi organisations are sceptical of the technology — which now has 25 million users a day — because of the security risks it poses. Having Microsoft’s security behind it will alleviate that fear, he said.
“They’re going to start running it on Teams pretty shortly so once that integration starts happening, that will allay some of the security concerns that organisations have.”
ChatGPT recently made headlines when it was found that users could put in the website address of news articles behind paywalls and the tool would return a regurgitated version of the article.
And nearly half of HR leaders are coming up with guidelines to regulate its use in the workplace, found a recent survey.
Confidentiality and bias in open-sourced AI
But Morrison said that the big thing for organisations to think about from a legal perspective is going to be confidentiality.
“Confidentiality will be a big one for anyone that’s holding people’s information,” said Morrison referencing the launch of Dropbox and confidentiality issues that arose from uploading sensitive information to a third party.
“It will be the same problem with ChatGPT. For example, if you put names into it, who knows if you’ve kept them confidential.”
Another pitfall of the technology that’s gained plenty of media attention is the potential for bias. USA Today, for example, said the chatbot showed anti-Conservative bias when asked about topics like affirmative action and diversity and transgender rights, which prompted a flurry of people to refer to the technology as “woke.”
Pros and cons to ChatGPT
Morrison agreed that the use of ChatGPT can lead to bad business practices and warned against employees being too reliant on ChatGPT.
“If you’re just putting everything into a chatbot and you’re not checking it, who knows what you’re saying? I’ve asked ChatGPT to give me a summary and it’s been wildly wrong, but if you didn’t know that, you could end up being way off the mark,” he said.
That was something the law industry found when it uncovered that ChatGPT creates case law that doesn’t exist.
“When you input ‘Find me a Commonwealth decision that deals with [insert issue here]’, the chatbot will return a case, but if you search for that case, then it returns no results,” said Morrison.
For all of its cons, the capabilities chatbots offer to organisations are huge. In the legal industry, there’s now a bot being developed where a lawyer can input all their evidence and the bot will tell you how likely you are to win, where the weaknesses are in your case, and how to go about fixing them.
“If you look at it broadly, it’s a good product, it’s a good start, said Morrison. “I can’t help but feel if you don’t use it, then you’re just swimming against the tide. It’s like anything new, you’re going to get an air of scepticism but if it’s faster and more efficient, people will use it.”
