Home OpenAI announces new ChatGPT guardrails

OpenAI announces new ChatGPT guardrails

April 26 — ChatGPT’s incredible capabilities can augment how we work, streamlining workflows and processes. Most business leaders are impressed with ChatGPT, but data privacy concerns have grown among CIOs and CISOs, leaving leaders with unanswered questions.

The forthcoming release of ChatGPT Business offers enhanced data privacy controls. Keep reading to see what’s on the horizon…

ChatGPT data privacy changes

  • Users of OpenAI’s ChatGPT tool can now disable chat history within the tool, meaning that personal information will not be stored. In addition, users can export ChatGPT data in order to analyze the information that the platform does store.
  • OpenAI is developing a new ChatGPT Business subscription to meet the needs of groups and enterprises seeking greater control over their data. With this new version, businesses will be able to manage end-users more effectively.
  • ChatGPT Business data policies are expected to mirror those of the ChatGPT API, where end-user data is not used to train models by default. OpenAI expects to release ChatGPT Business in coming months.

Further information

The new controls that OpenAI is rolling out today will be accessible to all users via ChatGPT’s settings toggle. “We hope this provides an easier way to manage your data than our existing opt-out processes,” wrote the company in a blog post.

In the business sphere, for CIOs and CISOs, data privacy within the ChatGPT tool is a top concern. The fear is that when employees input information into publicly available large language models, the data will be entered into a system designed to inform new responses – potentially compromising business data.

Recently, a Samsung Electronics employee reportedly entered sensitive data into ChatGPT, putting proprietary information at risk; possibly endangering business strategy.

Businesses, ChatGPT and Italy’s influence

In late March, the Italian Supervisory Authority imposed a temporary limitation that prohibited OpenAI from processing Italian users’ data. The regulatory body suspected that OpenAI was breaching the European Union’s General Data Protection Regulation.

Subsequently, Italy’s data protection watchdog published a list of demands for OpenAI. If those demands were met, the Supervisory Authority would lift the ban. One of the demands required OpenAI to add easily accessible tools that permit users and non-users to obtain their personal data. Another demand insisted that users must have the right to refuse access to their personal data for training the language learning model.

For more ChatGPT insights, please see CyberTalk.org’s past coverage. Want to stay up-to-date with trends in technology? Check out the CyberTalk.org newsletter. Sign up today to receive top-notch news articles, best practices and expert analyses; delivered straight to your inbox.