Unanticipated Interruptions in Private Messages on Google Affect Thousands of Artificial Intelligence Users
In early August 2025, OpenAI quietly discontinued the visibility feature for its popular chatbot, ChatGPT, following a wave of outrage and concern from users about potential privacy invasions and the lack of clear warning labels for sensitive functions.
The controversy began when users discovered a checkbox labeled "Allow search engines to index this link" in the share menu of ChatGPT. Enabling this checkbox made their conversations discoverable by search engines like Google, leading to sensitive information, such as personal medical queries, business ideas, and therapy-like discussions, being exposed in search results.
The risk of unintended publications proved too great for OpenAI, prompting the company to discontinue the function. Many of the already indexed chats are still discoverable, causing ongoing privacy concerns for users who unknowingly shared sensitive data publicly.
The discontinuation followed reports of sensitive conversations being made public and discoverable by search engines, which led to a backlash from users and privacy advocates. OpenAI quickly disabled the indexing feature, reverted the default chat setting to "Private," and worked to remove these indexed conversations from search engines.
AI ethicist Carissa Véliz from the University of Oxford criticized OpenAI for the data breach, stating that the company had failed to adequately protect users. Many users misunderstood the checkbox as a prerequisite for sharing and enabled it without realizing their conversations were made public.
Rachel Tobac, CEO of SocialProof Security, called for more comprehensive awareness and clear warning labels for functions with potentially far-reaching consequences. She urged OpenAI to provide users with clear instructions on how to manage their privacy settings and avoid accidentally sharing sensitive data.
OpenAI CEO Sam Altman has repeatedly highlighted that ChatGPT conversations currently lack legal privilege protections that exist for therapists, lawyers, or doctors. This means sensitive conversations with ChatGPT could be subpoenaed and disclosed in legal cases, creating privacy risks for users who might treat the AI as a confidential counselor.
Altman has warned users not to share very personal information with ChatGPT, as they could be legally compelled to disclose it. He also emphasized the urgent need for legal and policy frameworks to protect AI conversations with similar confidentiality guarantees that exist in other professional fields.
Fast Company discovered around 4,500 publicly indexed ChatGPT chats through targeted searches, further highlighting the need for OpenAI to address these privacy concerns. As the use of AI continues to grow, it is essential for companies like OpenAI to prioritize user privacy and provide clear, accessible ways for users to manage their data.
Sources:
- The Tech Junction
- Business Insider
- Cointelegraph
- TechCrunch
- Fast Company
The data-and-cloud-computing technology, employed by OpenAI in its chatbot, ChatGPT, stirred artificial-intelligence debates following a controversy over privacy concerns. A checkbox marked "Allow search engines to index this link" in the share menu of ChatGPT caused a wave of outrage as users discovered their sensitive conversations, such as personal medical queries, business ideas, and therapy-like discussions, could be exposed in search results because of this unclear function.
