A generative AI system (artificial intelligence software that answers questions), like ChatGPT and Bard, provides immediate responses to prompts based on the information it knows—much of the internet or a smaller content library if that’s what it’s been told to use. These responses usually look and sound authoritative—like a real human wrote them—although the responses are not always accurate. As a result, you may not be able to tell when an employee has incorporated AI-generated text into their work. Without realizing it, you may be relying on AI to make important decisions.
Are Your Employees Using Generative AI?
There’s a good chance you don’t know the answer to that question. An employee of yours may have recently used an AI tool to answer an email you sent them, provide information to a customer, conduct research, put together a proposal for stakeholders, or draft language to include as part of a product you sell.
Given all this, it would be prudent to create a use policy related to generative AI. There are benefits to using the technology, so an outright ban might not be in your best interest, but you should be aware of and take steps to mitigate the risks of using it. Here are some of the biggest risks you and your employees should keep in mind:
Generative AI can be wrong. Responses may be incomplete, contain factual errors or biases, and reference sources that don’t exist. If your employees use such a tool, they’ll need to vet the responses thoroughly. Otherwise, they risk providing incorrect information or sources to coworkers or customers who may then act on that faulty information.
Generative AI tools don’t necessarily tell you where they got their information. The people who originally came up with the ideas, insights, or guidance AI spits out don’t always get credit. Employees who use AI for work should take care that they’re not effectively plagiarizing the work of others—especially if their work will be visible to customers or the public.
Privacy and Security
In addition to being careful with the information they receive from generative AI, your employees should also take care with the information they provide it when asking a question. It should be a violation of company policy to use confidential information, trade secrets, and private company data in AI-generating tools.