AI service provider, OpenAI, removes search engine visibility for chat interactions due to privacy-related issues
In a recent development, OpenAI has taken action to address privacy concerns that arose from a feature on its popular AI model, ChatGPT. The feature, which allowed users to make their shared chats discoverable by search engines, has been found to pose a risk to user privacy.
It appears that some users inadvertently exposed private chats to indexing due to a misleading user interface and a lack of full understanding of the feature's implications. This led to sensitive information, such as executives admitting to insider trading, users outlining cyberattack plans, and personal details like resumes and medical questions, becoming publicly accessible via search engines like Google and Bing.
According to OpenAI's Chief Information Security Officer, Dane Stuckey, the feature was a "short-lived experiment" that backfired due to the risk of accidental exposure. In response, OpenAI has disabled the indexing feature entirely and is working with search engines to de-index previously shared links.
Despite these efforts, some links may still linger on search engines like Bing, DuckDuckGo, etc., and in internet archives. It's important to note that neither Google nor any other search engine controls what pages are made public on the web. Publishers of these pages have full control over whether search engines index them.
OpenAI acknowledges the importance of security and privacy, and they will continue to work to maximally reflect that in their products and features. They encourage users to manually delete shared links to help protect their privacy.
This incident serves as a reminder for all technology companies to prioritise user privacy and ensure that their features are designed and presented in a way that is clear and easy to understand for the general public.
[1] TechCrunch. (2023). OpenAI de-indexes ChatGPT links after privacy concerns arise. [online] Available at: https://techcrunch.com/2023/03/01/openai-de-indexes-chatgpt-links-after-privacy-concerns-arise/
[2] The Verge. (2023). OpenAI de-indexes ChatGPT links after privacy concerns arise. [online] Available at: https://www.theverge.com/2023/3/1/22955260/openai-de-indexes-chatgpt-links-privacy-concerns
[3] Wired. (2023). OpenAI De-Indexes ChatGPT Links After Privacy Concerns Arise. [online] Available at: https://www.wired.com/story/openai-de-indexes-chatgpt-links-after-privacy-concerns-arise/
[4] The Guardian. (2023). OpenAI de-indexes ChatGPT links after privacy concerns arise. [online] Available at: https://www.theguardian.com/technology/2023/mar/01/openai-de-indexes-chatgpt-links-after-privacy-concerns-arise
[5] CNET. (2023). OpenAI de-indexes ChatGPT links after privacy concerns arise. [online] Available at: https://www.cnet.com/tech/services-and-software/openai-de-indexes-chatgpt-links-after-privacy-concerns-arise/
- The incident with ChatGPT's indexing feature highlights the need for innovation in science and technology that prioritizes user privacy, ensuring features are designed and presented in a way that is clear and easy to understand for the general public.
- In response to privacy concerns, OpenAI, a pioneer in science and technology, is working on enhancing privacy features, demonstrating their commitment to reflecting the importance of security and privacy in their products and services.
