Since the early days of business, owners and operators have searched for ways to cut costs and increase profits in their industry via every possible avenue. Whether through a new product, a new invention, or new technology, business has always been about making ‘it’ better, cheaper, and faster. This overriding desire to stay ahead of the competition has led to countless revolutionary breakthroughs that have objectively been a net positive for the industry, and even the world. However, some corner-cutting measures have been more opaque in their benefits and it often takes years for the pros and cons of a new advancement to truly shake out. Enter ChatGPT.
Wave of the Future…?
The generative artificial intelligence tool has been turning heads in recent months for its ability to quickly create conversational language in a variety of styles and voices from brief, simple prompts. Taking this burgeoning technology a step further, Microsoft recently announced that ChatGPT’s A.I. features – known as Copilot in this particular context – will be integrated into the company’s most popular apps, including Word, PowerPoint, and Excel. Once implemented, the language model software is expected to save users hours of time by instantly analyzing their request and building content across a variety of platforms.
As expected, this is music to the ears of many owners, operators, CEOs, and other corporate managers whose businesses deal in content creation, whatever its form. The ability to potentially knock hours off every task brings with it the possibility of ramping up productivity to a much higher degree, and not just when it comes to direct content. Microsoft’s announcement confirmed that Copilot’s possible uses will be many, including presentations, emails, notes, and chats within Microsoft’s wide-ranging framework.
… or a Legal Minefield Waiting to Happen?
But not all is wine and roses when it comes to the near-limitless possibilities of ChatGPT’s A.I. power. The platform’s massive potential has given rise to a groundswell of worry over the applications of the technology and how it might be applied – or misapplied. Even Microsoft’s announcement admitted that ChatGPT is not infallible and should only be used as a ‘first draft’ when it comes to content creation, adding that “sometimes Copilot will be right, other times usefully wrong.”
Beyond whatever accuracy concerns there may be, industry insiders have also warned of other significant issues that could stem from employers directing their staff to use ChatGPT when performing their work tasks.
These include, but are not necessarily limited to:
- Knowledge gaps – ChatGPT’s database is vast, but not unlimited. Employees relying on ChatGPT for factual accuracy could be led astray by ChatGPT’s blind spots. There are also concerns regarding the sources of ChatGPT’s information. With so many disreputable fonts of data within the internet, insiders have openly worried that ChatGPT may not only have hidden gaps of knowledge, but some of its said knowledge may be questionable.
- Inherent bias – Earlier versions of ChatGPT’s A.I. model were lambasted by observers who noted that its use in recruiting was biased against women. Even when the offending data points were removed from the system, ChatGPT would re-learn the bias based on its pool of recruits. These earlier versions were eventually phased out due to compliance and discrimination concerns, but many worry that the problem is still lurking in ChatGPT’s code.
- Privacy concerns – An unexpected function of ChatGPT’s method of learning is its difficulty of partitioning and ‘saving’ knowledge that should not be shared. Although Microsoft has assured users in the past that nothing that is told to the program is “saved,” per se, it does “learn” from everything it is told. Insiders worry that proprietary information, should it be shared with ChatGPT, could eventually be regurgitated by the program later in a way that could violate the secrecy of said information.
ChatGPT’s immense time-saving power may seem like an irresistible tool by which employers can maximize efficiency and increase their content capacity. But with these apparent rewards come the potential risk that the tool’s possible weak points regarding confidentiality, bias, and accuracy may outweigh its usefulness. The time is now for employers to consider whether ChatGPT can be safely utilized within their companies and how its usage should be monitored and implemented to avoid future potential legal pitfalls.
This is just one of many examples that show how complex and fraught with danger the world of business law can be. The safe and proper operation and governance of your business is often more complex than it first appears, and new inventions, technologies, and ideas are forever testing the boundaries of what business regulations allow. If you are in need of trusted and experienced counsel in the realm of business law, Richards Rodriguez & Skeith’s business and transactional attorneys may be able to help. Contact us today!