In a blog post released this week, OpenAI revealed that it would give users the ability to turn off chat history functions for the generative AI system. The Microsoft-backed company hopes the move will go some way to responding to experts’ criticism of the platforms data security credentials
ChatGPT has released a new “history off” mode, enabling users to decide which conversations with the AI platform are used in the technology’s development models.
In a blog post released this week, OpenAI revealed that it would give users the ability to turn off chat history functions for the generative AI system. The Microsoft-backed company hopes the move will go some way to responding to experts’ criticism of the platforms data security credentials.
Conversations won’t be used by OpenAI to train the system’s underlying models when the “history disabled” option is ticked, and relevant chats will not appear in the programme’s sidebar. Dialogues will be held on the organisation’s central servers, but will only be looked at if abuse checks need to be made, and they will be erased after 30 days.
“These controls, which are rolling out to all users starting today, can be found in ChatGPT’s settings and can be changed at any time. We hope this provides an easier way to manage your data than our existing opt-out process. When chat history is disabled, we will retain new conversations for 30 days and review them only when needed to monitor for abuse, before permanently deleting,” the blog post reads.
OpenAI has also announced the development of a new subscription service called ChatGPT Business, aimed at professionals and enterprises looking for greater control over their data. According to the company, ChatGPT Business will adhere to the API’s data usage policies, which means that user data will not be used to train their models by default. The subscription service is expected to be available in the coming months.
In addition, a new “Export” option in the settings will allow users to easily export their ChatGPT data and understand the information that the platform stores. Users will receive an email containing a file with their conversations and all other relevant data. The new feature is expected to simplify data management for ChatGPT users.
ChatGPT has been facing regulatory hurdles in Europe over its compliance with the region’s data protection laws. The European Data Protection Board initiated an investigation into ChatGPT, following complaints from privacy watchdogs that the chatbot failed to adhere to the EU’s General Data Protection Regulation.
This comes after Italy temporarily banned the use of ChatGPT in March, alleging violations of user privacy. The country’s data protection guarantor had demanded that the service prove its compliance with relevant privacy laws and enhance transparency around user data management.
According to Nader Henein, the Vice President and Analyst at Gartner, privacy and data governance were not the top priorities for OpenAI at the start of their venture. However, he added that such oversight is not unusual for a start-up that is primarily focused on delivering a functional product to the market.
PrivSec Global brings together leading experts from around the globe, for a 2-day livestream experience that ensures attendees have access to the latest information, guidance and advice on data protection, privacy and security.
PrivSec Global returns on 17th & 18th May 2023, and will once again deliver a carefully curated agenda that taps into the expertise of subject matter experts, industry leaders and academics.
→ Generative AI and the GDPR: A Match Made in Hell?
- Day 2: Thursday 18th May 2023
- 12:30 - 13:15
- Dr Pieter van der Walt, Group Chief Privacy Officer, Discovery LTD
- Cristiana Deca, CEO & Data Protection Expert, Decalex Digital
- Huzefa Tavawalla, Head: Bangalore Office & Disruptive Technologies Practice Group, Nishith Desai Associates
- Dr. Nathalie Moreno, Partner, Addleshaw Goddard LLP
No comments yet