Attorney
The Privacy Predicament: The Risks of Generative AI
What is ChatGPT and Generative AI?
Since its introduction in late 2022, there has been much discussion regarding AI chatbots, with ChatGPT at the forefront. But what is ChatGPT and generative AI? In short, ChatGPT is a natural language processing tool powered by AI; in practice, the platform responds to user prompts by utilizing machine learning and statistical data to arrange words in the order in which they are most likely to appear.
What are the applications of ChatGPT?
Given its diverse application, people have used ChatGPT for many purposes, such as engaging in human-like dialogues, organizing data, and creating recipes. Since ChatGPT’s introduction, a question has remained: What are the risks of this technology?
What are the risks associated with Generative AI tools like ChatGPT?
While copyright commands the most people’s legal concerns with Generative AI tools like ChatGPT, privacy is a lesser-known but equally troubling issue. With privacy, there are two main concerns you should familiarize yourself with. The first is that as a provider of any Generative AI services, the program may inadvertently reveal sensitive or personal information.
On the flip side – for users – there are potential risks that ChatGPT could utilize confidential or sensitive information for other unexpected purposes. For example, someone may turn to AI to seek a diagnosis for medical symptoms, exposing sensitive health information; they could input financial records for data organization or even use its translating ability on a private message; all of these scenarios could expose users to undue risk as user provided data is used to train the platform for future interactions.
OpenAI’s March 2023 data breach
While OpenAI, the company responsible for ChatGPT, assures that it will never utilize personal information to create profiles, contact people, or advertise to them, the platform contains vast data, making it an attractive target for malicious actors to exploit. In March 2023, OpenAI reported a data breach in which users’ names, conversation histories, email addresses, and more were leaked (LINK).
Therein lies the danger; since it’s such a young technology, we don’t know what’s in peril. The hope is that the California Consumer Privacy Act (CCPA) and General Data Protection Regulation (GDPR) may provide some relief. Only time will tell.
What Steps can you take to stay safe while using ChatGPT?
What can you do to stay safe? There are a few basic steps users can take: as with most internet platforms, it’s important to maintain data confidentiality and privacy, be careful with what information you provide to the website, monitor your usage, check chatlogs for unusual activity, and be responsible with how you make use of this new technology.