5 ChatGPT Mistakes To Avoid in the Workplace

Updated June 22, 2023

A person types on their laptop at a large table in an office setting.

You’ve probably heard of ChatGPT, the advanced chatbot that’s changing the way we look for jobs and even do our jobs more efficiently. Chances are, you’ve probably tested out ChatGPT to see what it can do for you, whether it’s to write an itinerary for an upcoming vacation, outline a presentation or draft code.

Despite the vast capabilities of ChatGPT, it’s imperative that we also discuss its limitations in the workplace. It’s still not a perfect technology. In fact, many companies are pumping the brakes on ChatGPT and other generative AI software for fear of misuse.

In the first half of 2023, a slew of companies have placed restrictions or complete bans on ChatGPT. The primary concerns center on data privacy, as many companies work with sensitive information and proprietary code.

Samsung: Fears became a reality at the electronics company Samsung when staff accidentally leaked sensitive data via ChatGPT. This resulted in a company-wide restriction of all generative artificial intelligence tools on company-owned devices.
Verizon: Similarly, Verizon also banned the software on company devices to prevent loss of customer information and source code.
Wall Street: Investment banks including Citigroup and Deutsche Bank have all banned usage of ChatGPT over concerns about private financial data being shared with the chatbot.

We would be remiss to not mention Italy. It’s not a company, but this country issued a temporary ban on ChatGPT around privacy concerns. OpenAI made swift updates to meet regulatory requirements and reverse the ban, but this is the first example of a government taking action against ChatGPT. This situation may foreshadow future conversations around generative AI and data protection laws.

5 ChatGPT mistakes to avoid at work

Even if your company hasn’t outright banned ChatGPT, it’s critical to know the common mistakes that can put your job or professional reputation at risk. The following is a list of what you should avoid when attempting to utilize ChatGPT for workplace assignments, as well as possible solutions to help you avoid each potential pitfall.

Mistake #1: Assuming your confidential data is safe

Anything you put into ChatGPT is presumably used by OpenAI to continually improve the software. Also, the platform is vulnerable to bugs, like what occurred with the March 20th outage. This bug in an open-source library allowed some users to see the titles from other users’ chat history. Additionally, the bug exposed some users’ payment-related information.

This bug was relatively small in the grand scheme of things, but employees working with sensitive information should lead with caution. Threat actors have increasingly targeted open-source libraries, with attacks increasing by 742% since 2019.

In order to keep you and your work safe when using ChatGPT, think twice before inputting any personal information into it, such as your full name, company name, phone number or address. Refrain from using any NSFW language or using ChatGPT for NSFW purposes. If you’re using ChatGPT for work purposes, consider asking your boss or HR manager to provide you with a temporary company card or highly-protected payment platform to engage in low-risk transactions.

Furthermore, consider working on a VPN-enabled device to add another layer of overall data protection. Ensure that your browser firewall is up and maintained, and the same goes for the firewall on your device. Double-check that your virus protection software is enabled, and clear your cache regularly to clear any potential for viruses. Avoid logging into suspicious sites or opening suspicious emails, and steer clear of public Wi-Fi portals. If you run into any problems, contact your IT department immediately.

Mistake #2: Believing everything ChatGPT says is true

You only need to spend a little time using ChatGPT to discover that it’s not the omniscient supercomputer that some may assume it is. Quite the opposite: the software has been known to generate comedically false information, produce biased or discriminatory content and even hallucinate.

This is why it is imperative to carefully edit and cross-reference everything ChatGPT writes—or your boss might start wondering if you’re hallucinating, too. After receiving data from ChatGPT, be sure to cross-check the information using other online tools, such as plagiarism detectors. List each claim within the text and do a quick search to double-check the information, being sure to only use trusted sources.

Peer reviews can also prove to be an extremely effective way to help weed out any false information, so don’t be afraid to ask a co-worker or your boss to provide an extra set of eyes. Considering the negative effect a false claim may have on your professional reputation, take your time when reviewing and be as thorough as possible.

Mistake #3: Letting ChatGPT do your job for you.

It’s important to remember that generative AI like ChatGPT operates like a search engine. Therefore, it’s entirely reliant on data it can access and will regurgitate this existing information instead of generating something new.

That’s why ChatGPT is intended to be used as a tool—not a stand-in for a human employee. It can’t do critical thinking, ethical decision-making or come up with ideas that are truly innovative or original. That’s still your job!

Rather than immediately jumping into ChatGPT for assistance on an assignment or project, do some quick research on the limitations of generative AI to better understand what it may or may not be able to accomplish for you. Search up ChatGPT best practices to ensure you’re optimizing your time and effort, and never submit ChatGPT data without reviewing it thoroughly.

Mistake #4: Relying on ChatGPT instead of upskilling.

Becoming too reliant on ChatGPT can present a number of problems. The software might not always be available due to technical issues or bugs. As previously discussed, your company might decide to ban the usage of ChatGPT altogether.

Depending on ChatGPT to do aspects of your job such as drafting emails, responding to customers or structuring information, may prevent you from developing these important skills on your own. Rather than trying to utilize ChatGPT as a problem-solver, consider only using ChatGPT to brainstorm solutions or consult on pathways, thereby making you that much more effective and efficient when working towards a solution.

Mistake #5: Replacing face-to-face interaction with ChatGPT.

There are some job responsibilities that simply cannot be replaced by any AI. These include situations where soft skills are key, such as helping to train someone to work at your company, managing a conflict or addressing a client concern. ChatGPT cannot replicate empathy, personal connection, eye contact or a firm handshake.

With all of this in consideration, ChatGPT is still one of the most talked-about technologies in the work world today and companies are certainly taking notice. According to a survey from Resume Builder, nine in ten hiring companies want workers with ChatGPT experience.

As some companies are completely banning the software, others have embraced it for improving efficiency and freeing up workers for other high-priority tasks.

Miami-based hedge fund Citadel is working toward securing a company-wide ChatGPT license.
Shopify is using ChatGPT to help with user inquiries and provide tailored recommendations.
Instacart is experimenting with ChatGPT for providing recipe recommendations and answering food-related questions.
The New York City school district reversed a ban on ChatGPT, with Chancellor David Banks writing, “While initial caution was justified, it has now evolved into an exploration and careful examination of this new technology’s power and risks.”

While exciting new technologies like ChatGPT bring great potential, it is imperative for employees and employers to approach it with caution and thoughtful consideration to mitigate unintended consequences.