Uh Oh! ChatGPT Conversations Now Searchable on Google
In late July 2025, it was revealed that more than 4,500 shared ChatGPT conversations had been indexed by Google and were publicly searchable. These conversations included personal names, emails, sensitive disclosures, and in some cases, proprietary business information.
The root cause was linked to users sharing chats via the "Share" feature in ChatGPT and selecting the "Make this chat discoverable" option. Those links were then crawled and indexed by search engines.
In response to privacy concerns, OpenAI removed the “discoverable” option on August 1, 2025, and began working with search engines to remove previously indexed content. However, some of this information may still be available via third-party archives such as the Wayback Machine.
LLMs like ChatGPT continue to be valuable tools for business and creative work. They can accelerate brainstorming, help generate high-quality drafts quickly, and serve as powerful collaborators in ideation and content creation. For small and mid-sized companies in particular, they offer a flexible, low-cost way to increase productivity and enhance innovation.
The recent exposure of ChatGPT conversations offers a clear lesson: conversations shared publicly are not private. Users assumed their shared links were confidential unless posted somewhere intentionally. In reality, any link marked as "discoverable" became visible to search engines and open to the internet.
Here are key recommendations for teams and organizations:
Search Google to check if any past shared conversations are publicly viewable.
Audit how AI tools are being used internally, particularly for content involving client data, company strategy, or proprietary intellectual property.
Educate staff to treat chatbot outputs like emails or cloud documents. If it's not something you would publish on a public website, it should not be shared through AI tools without careful review.
Consider disabling or limiting share features in enterprise AI platforms. For sensitive use cases, evaluate private deployment options or platforms with more robust data governance.
LLMs offer enormous potential for business growth and innovation. But like any powerful technology, they require thoughtful use. The ChatGPT incident shows how easily well-intended use can lead to unintended exposure. At NOV8TIV, we help clients take full advantage of emerging technologies while safeguarding the integrity of their operations and intellectual property. If your organization uses AI for internal work, now is the time to revisit your policies and make sure you are protecting what matters most.