Artificial intelligence, including so-called “large language models” like ChatGPT, has rapidly become a talking point in the press, amongst governments, and maybe even in your office. While AI has been a background subject for decades, everyday web users can now engage with AI like never before.
But, whenever there’s a technology change, it’s always wise to think about the security issues. With any shiny new technology, you should consider security and privacy risks before diving in. When it comes to AI-powered language models and other services, there are five major factors to consider when loading up AI for help at work, school, or for fun:
1. Don’t hand over personal information
AI models partly learn from what users input into the system. Therefore, you shouldn’t input any information you want to keep private, from your company’s proprietary computer code to sensitive information about yourself or your family, like addresses, account numbers, or passwords.
2. Prompting isn’t the same as creating
When it comes to your child’s homework or your work endeavors, know that putting a query to AI and copying/pasting the results isn’t the same as doing the work yourself. Also, if you’re asking a fact-based question, fact check everything. These models have become infamous for giving very confident but wrong information in many situations. Other times, people have noted that AI models produced bizarre—and sometimes creepy—responses suggesting that the model had a mind of its own, which have been deemed “hallucinations.” It’s best to look at them as tools: they can help you get the work done, but you’re more talented than a machine!
3. Privacy concerns
There are many concerns over how AI models scrape the web, from how these programs use artists’ and writers’ creations to what personal information they know about us. Many experts worry that it’s collecting data on children, for example, and how these services can alert people about sharing their data remains an open question. In many cases, your chats with an AI aren’t private—the company can see what you input, even if it’s anonymized. Carefully read the privacy notices of any AI service you use and ensure that you’re okay with sharing the data it collects.
4. Cybercriminals also use AI
Another trend is the rise of cybercriminals using AI to get better at their crimes. There’s evidence that bad actors are using AI to craft more deceptive phishing emails and help develop malware. When there’s any big disruption in tech, take it as a good time to review your cybersecurity basics: use strong passwords, take advantage of password managers, and enable MFA for all accounts that allow it.
5. Read your company policy
As AI has risen in popularity, many companies have enacted guidelines on its usage in the workplace. For example, some companies might require their employees to disclose the use of AI in their work. Others might require those who use AI to receive training on how it works, as well as how to use it ethically. Some companies may prohibit the use of AI altogether. Regardless of what your stance is, be sure to stay informed of your company’s AI policy.
AI can be extremely helpful for everyday situations, whether you’re using it to brainstorm ideas, summarize long articles, or even create an itinerary for an upcoming trip. But, like with any new technology, it’s important to be aware of security and privacy risks before fully immersing yourself. By keeping the above tips in mind, you can safely incorporate AI into your life and keep your online presence secure.
