News / Legal Brief
The ‘AI Chatbot Race’
Feb 20,2023
By Kyra South, Senior Associate and Janice Geel, Candidate Attorney, reviewed by Natalie Scott, Director, Head of Sustainability Practice
2023 will be remembered as the year that Artificial Intelligence took the world by a storm. In just two months, ChatGPT has garnered 100 million users, a feat that took TikTok nine months and Instagram two years to achieve.[1] Hot on ChatGPT’s footsteps are Google’s “Bard“[2] and Microsoft’s “New Bing“[3], both of which are due to be released as rivals to ChatGPT in what is being dubbed the ‘AI Chatbot Race’.[4]
In our article titled “ChatGPT: magic bullet or the beginning of the end?“[5] published on 30 January 2023, we noted that legal practitioners could use technology such as ChatGPT to enhance the quality of their work (if used correctly) but that in the absence of a regulatory framework,[6] the legal practitioner is responsible for verifying the information received from ChatGPT with multiple other sources.
In and amongst the excitement and eagerness that is currently gripping Artificial Intelligence chatbot (“AI Chatbot“) users, AI Chatbot users (including legal practitioners) must be cognisant that the information generated by an AI Chatbot may be inaccurate, incorrect and misleading, which leads to the following question:
If a legal practitioner uses and relies on information delivered by an AI Chatbot, and which information is subsequently discovered to be incorrect, who would bear the ultimate responsibility for damages suffered and losses incurred due to such information being incorrect?
In addition to the questions and opaqueness surrounding ultimate responsibility, legal practitioners must also be cognisant of South Africa’s data protection laws[7] when researching a new topic through AI Chatbots and be mindful not to inadvertently publish a client’s Personal Information[8] through the AI Chatbot, unless that Personal Information is (a) published in compliance with the Protection of Personal Information Act 4 of 2013 or (b) publicly available.
ChatGPT (for example) reminds users not to share any sensitive information[9] with the chatbot when inputting queries as each input is permanently recorded on ChatGPT’s database and is not capable of being erased.
On Wednesday 8 February 2023, the world witnessed the consequences of an incorrect response, when Google’s holding company, Alphabet Inc. lost US$100 billion (approximately 8% of its stock price[10]). Bard (Google’s AI Chatbot) incorrectly stated that the James Webb Space Telescope “took the very first pictures of a planet outside of our solar system“,[11] a fact easily verifiable, and which highlights that (i) responses generated by AI Chatbots can be wrong, (ii) information generated by AI Chatbots must be verified with multiple other sources and (iii) relying on incorrect information can have serious consequences, including significant financial implications.
“The AI Chatbot Race
In a digital world of code and grace,
A race has begun, at a breakneck pace,
Where chatbots compete, with wit and flair,
To become the best, beyond compare.
The finish line’s in sight, with victory near,
As AI algorithms are engineered with care,
To deliver responses that are sharp and bright,
And keep users engaged, both day and night.
Some chatbots are simple, with just a few lines,
While others are complex, with intricate designs,
They use NLP and ML, to understand and reply,
To questions and prompts, that they can’t deny.
The race is on, with no time to spare,
As each chatbot strives, to be beyond compare,
With accuracy and speed, as their hallmark,
In the quest for excellence, they give it their all.
The winner takes all, in this digital space,
With accolades and fame, for their tech race,
And as the world turns, and the race does end,
The best AI chatbot will rise and transcend.
So let the race begin, and may the best one win,
In this digital arena, where AI does begin,
To shape our future, with its stunning might,
In this AI Chatbot race, thats taking flight” [12]
