top of page

ChatGPT and Privacy Concerns

ChatGPT has taken the large language space model by storm and is one of the biggest technology revelations of our times. Within two months of its launch in January 2022, ChatGPT had 100 million active users. It took TikTok nine months, Google one year, Instagram over two years, Snapchat over three years, and Facebook over four years to reach this user mark. The advanced conversational AI chatbot created by OpenAI is designed to answer questions, provide humanlike AI interactions, and solve a range of problems. Such is the popularity of ChatGPT, Google has also unveiled the launch of its conversational AI called Bard.

While the platform has been applauded for this innovation, it also brings serious privacy concerns and fear of violation/misuse of proprietary and private information. Some of the crucial problems envisaged as an implication of such new and innovative AI are:

  • ChatGPT has been trained and fed massive amounts of data and human text scraped from all over the internet. The data can refer to actual individuals and things that they have written or said over some time in varied contexts and analyze them in a manner that is not apparent and also, at times, inaccurate or irrelevant to the context. Moreover, the information can be proprietary or copyrighted. For example, scraping data points from sites in the EU can breach the GDPR.

  • OpenAI has never taken any consent/permission for the use of such data, which can be related to the personal information of individuals and is a clear violation of privacy, especially when the data is sensitive personal information about individuals, which ChatGPT can reveal to users whenever the query/context requires. It does not appear sufficiently transparent; it might be gathering and improperly using personal data, and people would have trouble exercising their right to manage their personal information.

  • Considering that OpenAI was recently valued at USD 29 billion, it is surprising that the businesses, websites, or people whose data ChatGPT exploited without permission have not received compensation from OpenAI.

  • ChatGPT breaches the contextual integrity of the information that is publicly available. The principle requires that the information not be used or shared outside its original context.

  • OpenAI does not provide any methods for people to review or delete their personal information as required by numerous privacy laws like the CCPA and GDPR. However, the problem with natural language processing tools like ChatGPT is that extracting or deleting any individual’s data or even traces of an individual’s personal information with such pre-existing data already fed into the systems is impossible. Whether ChatGPT is compliant under various privacy regulations or not is still debatable.

  • As an open tool, ChatGPT can be abused by bad actors to carry out phishing and scams asking victims to divulge sensitive information through realistic-sounding conversations and expert impersonation. Also, it will be simple for such individuals to disseminate false material that sounds quite credible.

This is what a ChatGPT application had to say about ‘Privacy concerns with ChatGPT’:

“ChatGPT presents several privacy concerns. ChatGPT systems store user data, including personal information such as messages sent, contacts listed, and even conversations with the AI. This data may be accessible to third-party companies and could potentially be used for malicious purposes. Additionally, the system may be vulnerable to malicious attacks, and compromising user data could occur due to poor security measures. Additionally, the AI may be programmed to access user data without the user’s knowledge and permission, which could lead to a violation of privacy rights.”




Follow LexTalk World for more news and updates from International Legal Industry.




bottom of page