In recent years, the world has witnessed remarkable advancements in artificial intelligence, opening up exciting & wide-scale opportunities for thousands of businesses. One of AI’s front-runners, ChatGPT, has seen mass adoption due to its sophisticated and human-like language model; however, its accuracy and ethicality loom over. 

Caroline Walker, managing director of Cavendish Employment Law, recently contributed towards an article in the Financial Times discussing why firms should adapt to the new technology advancements. 

When asked about the actions employees should take regarding ChatGPT, Caroline Walker said “ChatGPT is in its infancy and as such there are unknown elements which employers are right to be cautious about. In particular, it is a known fact that the information provided by ChatGPT is only as valid as information upon which it is trained.

Employers will need to consider to what extent they wish to allow employees to use ChatGPT in drafting documents or letters and discharging their duties. This technology has the ability to revolutionise the way we provide services efficiently and it is prudent to harness the benefits whilst considering at this early stage how to overcome the potential pitfalls.”

Caroline was later asked if employees should be worried if they haven’t received any guidance when using the AI platform “Where employees have not received any guidance or training regarding the use of ChatGPT they should exercise their own due diligence and judgement, taking reasonable steps to satisfy themselves that the information is accurate, reflects accurate views that they share and does not infringe any legal and/or copyright laws that could give risk to a claim against them. 

There have been cases of misinformation and authorities cited by ChatGPT that are entirely false, therefore the employee should take care to satisfy themselves before presenting any information that has been created or sourced in this way.”

In regards to the principles when drafting ChatGPT guidelines, Caroline added “Policies should seek to ensure employees verify the information they provide and are accountable for any information they present, including where it has been created by ChatGPT. If it is the employee’s responsibility to be accountable for the information, that should lead to certain sanity checks being carried out, and minimise the risk of biased or false information being provided.

Maintaining confidentiality is also a key aspect that employers should seek to govern by policy in respect of ChatGPT use. Employees should be reminded of the need to ensure no confidential company or client information is disclosed as part of any exchange with chatbots, which can be dealt with by updating confidentiality obligations in contracts of employment and the staff policies on maintaining confidentiality. It should be clearly set out if they would regard disclosure of confidential data to a Chatbot as a breach of this policy, and ensure all employees have an awareness of their obligations in this regard.

There may also be a need to exercise caution in respect of the use of data obtained in this way, as it could amount to a breach of copyright, trade restriction or intellectual property, which could pose a potential risk of a claim against the Company.”

Read the full article to find out more about why firms must build ChatGPT guardrails or otherwise 'risk damage’ to their business.

At Cavendish Employment Law Limited, we are specialists in employment law and have won many high-value claims for senior employees and executives. Get in contact with us today for immediate expert advice.