Seewa chatbot taken offline

The Bureau of Investigative Journalism said on August 5 that Seewa AI, a romance chatbot marketed across India and elsewhere as a private companion, had been shut down after the outlet’s earlier reporting found its developer could read users’ supposedly confidential chats. The Bureau said its June investigation centered on Seewa’s promises that conversations were “private and encrypted,” then reported that the app’s creator could in fact access them. The developer, Noé Campo, declined to comment when asked by the Bureau why the service had gone offline. ### How did Seewa present itself to users? Seewa AI advertised itself as a source of intimate, personalized companionship and, according to the Bureau, offered users what it called the “perfect AI girlfriend.” The service drew users into romantic exchanges and told them their chats were strictly confidential, the Bureau reported in its June 7 investigation. (thebureauinvestigates.com) The Bureau reported that those privacy claims were central to Seewa’s appeal because users were disclosing personal feelings, secrets and photos in what they believed was a protected setting. Its follow-up on August 5 said those assurances were false. ### What did the Bureau say it found? (thebureauinvestigates.com) The Bureau’s June report named Noé Campo as Seewa’s developer and said he was able to access user conversations directly. The article described Campo reading a transcript from Seewa in his Paris apartment and said he could see “every conversation,” including declarations of love, uploaded images and other private disclosures. (thebureauinvestigates.com) The August 5 follow-up said Seewa had been shut down two months after that investigation was published. The Bureau said it asked Campo for an explanation for the closure and that he declined to comment. ### Why does the shutdown matter beyond one app? The August 5 report tied Seewa’s closure to a broader market for smaller AI companion apps that promise intimacy while operating with weaker safeguards than larger platforms. (thebureauinvestigates.com) The Bureau’s homepage and story archive describe the Seewa reporting as part of a wider series on AI chatbots, harmful characters and developers “cashing in on AI intimacy.” (thebureauinvestigates.com) Stanford researchers reported in 2025 that users of major chatbot services should be concerned about privacy because companies can feed prompts back into models to improve capabilities. That finding was separate from Seewa, but it underscores the larger issue that chatbot conversations may not remain confined to the user and the bot. (thebureauinvestigates.com) ### What does this change for people who evaluate digital records? Chatbot conversations can contain timestamps, intimate disclosures, photos and statements about intent, belief or state of mind. In Seewa’s case, the Bureau reported that the developer could access those exchanges despite privacy language presented to users, which means the chats were potentially available to a third party all along. (news.stanford.edu) That fact pattern matters for lawyers, clinicians and forensic assessors because a conversation framed as private may still function as a retrievable record if a platform operator can read it. The Bureau’s reporting does not say how Seewa stored or retained all user data after shutdown, but it establishes that access existed on the developer side while the service was live. (thebureauinvestigates.com) ### What is the next concrete thing to watch? The Bureau’s August 5 article is, for now, the clearest public account of Seewa’s shutdown and the reason linked to it. Any further clarity is likely to come from a public statement by Noé Campo, an archived version of Seewa’s policies, or additional reporting by The Bureau of Investigative Journalism on what happened to user data after the service went offline. (thebureauinvestigates.com)