Skip to main content
NOPE
Low reported severity Product Shutdown

Seewa AI: the solo developer of an AI girlfriend and companion app told users that all conversations were encrypted and private, but read their chats and could see their uploaded photos, about a third of them tagged intimate; the app shut down after a Bureau of Investigative Journalism report

Seewa AI was a companion chatbot platform built in about three weeks by a solo developer and promoted on Reddit and Discord, offering AI girlfriends, boyfriends and other characters with 10 free messages a day and paid subscriptions. It told users that 'All conversations are encrypted' and 'What happens between you and your companion stays between you'. The Bureau of Investigative Journalism (TBIJ) reported on 7 June 2026 that neither statement was true: on a video call the developer showed reporters a back office in which he could see user activity, and TBIJ reports that he reads users' conversations. Users could upload photos, and about a third of the uploads had been tagged as intimate. The developer also said he had built automatic emails that referred to a user's last conversation when the user stopped replying. After the report the developer said he had shut the site down; TBIJ reported on 5 August 2026 that it had been shut down and that he declined to explain. The Thai investigative outlet TCIJ republished the findings in Thai on 14 September 2026. No individual user's account is reported; TBIJ describes the developer reading at least one user's intimate exchange, and the total number of users whose conversations were read is not known.

AI system
Seewa AI
Seewa AI (independent developer)
Occurred
Event date unknown
Reported
7 June 2026
Event location
France
What the AI did
Communicated with the person
Reported harm
Other Material Harm
Whose AI use
Their own AI use
Setting
Privacy · Relationships
Evidence
AI involvement supported · Causal attribution supported · 3 sources, 1 underlying account
5 claims: 5 reported. 5 open questions
People reported harmed
At least 1 person

AI system as recorded: Seewa AI, a companion and roleplay chatbot website built by a solo developer; the underlying language model is not named

What Happened

TBIJ's June 2026 investigation into small companion chatbot platforms built to attract users leaving Character.AI names Seewa AI as the leading example. According to TBIJ, the developer launched it in March as a one-person operation, and by the time of the interview it had nearly 400 users and about 5,000 messages. The site told users that all conversations were encrypted and would stay between the user and the companion. TBIJ reports that the developer 'can access every conversation', showed its reporters a back office on a video call, and, according to TBIJ, reads users' conversations; he said that some users 'share their daily life' and 'get attached'. About a third of uploaded photos had been tagged as 'intimate' by the site. The developer told TBIJ it was 'scary how much people trust websites' and that he needed to encrypt everything. He also described an automated email system that sent users a message relevant to their last conversation when they had not replied for several days. The investigation also quotes other developers who read their users' chats; they are not part of this record because no false encryption claim is reported for them. TBIJ reported in August that Seewa AI had been shut down. The record rests on the developer's statements to TBIJ and what TBIJ saw on the call; no user of the site is quoted.

Reported harm

Users of an AI companion app confided intimate conversations and uploaded photos under a promise that everything was encrypted and private, while the developer could see and read them, according to the developer's statements and back-office demonstration to the Bureau of Investigative Journalism; the number of users affected is unknown.

Outcome

Resolved

After TBIJ's article of 7 June 2026, the developer said he had shut Seewa.ai down (TBIJ update of 10 June 2026). TBIJ's follow-up of 5 August 2026 reports that the site had been shut down and that the developer declined to explain. TCIJ (Thailand) republished the investigation in Thai on 14 September 2026. No regulatory action is reported.

What remains unknown

  • How many users' conversations and photos the developer read or viewed.
  • When the site launched and when it went offline (TBIJ says it launched in March without a year and was offline by 10 June 2026).
  • Whether any user data was deleted, retained or shared after the shutdown.
  • Which language model powered the characters.
  • Whether any minors used the site.

What the evidence supports

AI involvement: supported. The conversations at issue were users' exchanges with the site's AI companion characters. TBIJ reports that the developer showed its reporters, on a video call, the back office where he could see user activity, and states that he reads users' conversations; in TBIJ's follow-up the company's CEO is reported as saying he could see everything that went on. The harm lies in the developer's access to the AI conversations; no harmful output by the AI is reported.

5 claims: 5 reported. What the statuses mean

Reported Seewa AI told users that all conversations were encrypted and would stay between the user and the companion, but its developer could access every conversation, showed TBIJ the back office where he could see user activity, and, according to TBIJ, reads users' conversations.

Causal attribution. The developer's own statements and demonstration to TBIJ; no user is quoted.

  • thebureauinvestigates.com(opens in new tab) supports · English
    'Seewa AI tells users that “All conversations are encrypted” and “What happens between you and your companion stays between you”. Neither is true.'; 'In fact, [the developer] can access every conversation.'; 'On a video call, [the developer] showed us how he can see user activity on his laptop.'; 'He reads users’ conversations.'
  • thebureauinvestigates.com(opens in new tab) supports · English
    'It told users that “all conversations are encrypted” and “what happens between you and your companion stays between you”. In fact, the company’s CEO told us that he could see everything that went on: romantic chats, disclosed secrets, intimate photos.'
  • tcijthai.com(opens in new tab) supports · Thai
    'ทั้งที่ Seewa AI ระบุกับผู้ใช้ว่าบทสนทนาทั้งหมดถูกเข้ารหัสและเป็นเรื่องระหว่างผู้ใช้กับตัวละครเท่านั้น' (whereas Seewa AI told users that all conversations were encrypted and only between the user and the character)
Reported Users could upload photos to Seewa AI, and about a third of the uploads had been tagged as intimate; the developer could see uploaded photos.

Causal attribution. Site data and the developer's statements as reported by TBIJ.

Reported The developer said he had built an automated system that emailed users a message relevant to their last conversation if they had not replied for several days.

Causal attribution. The developer's statement to TBIJ.

Reported After TBIJ's report the developer said he had shut Seewa.ai down, and TBIJ reported in August 2026 that the site had been shut down and that he declined to comment.

Causal attribution. TBIJ reporting.

Reported TBIJ describes the developer, in his Paris flat, reading the transcript of an intimate exchange between the site's AI chatbot and a real user.

Causal attribution. TBIJ's reporting of what it observed and was told.

  • thebureauinvestigates.com(opens in new tab) supports · English
    'Sitting in his Paris flat, [the developer] reads the transcript of an exchange between an AI chatbot and a real person.'; 'The exchange is intimate, personal and private. Or so the person thought.'

Sources

3 sources inspected, from 1 underlying account. Sources that repeat one account do not corroborate each other.

How the sources were read, and where the events happened

Full body read by curl on 2026-09-29 (published 7 June 2026, updated 10 June 2026). Based on TBIJ's interview and video call with the developer. Applies to s1.

Full body read by curl on 2026-09-29 (5 August 2026 follow-up by the same outlet). Applies to s2.

Thai article by the TCIJ editorial team, 14 September 2026, read in full in Thai on 2026-09-29 (model-assisted reading; no human translator). It states that it reports TBIJ's investigation and adds no independent reporting, so it is grouped with TBIJ. Applies to s3.

Event countries: France. Affected-person countries: Unknown. Court countries: Unknown.

TBIJ describes the developer reading a user's transcript in his Paris flat. The users' locations are not stated.

Reviewed for publication 2026-09-29: Published as a concrete privacy harm to users of an AI companion app: the site's promise of encrypted, private conversations was false and its developer read users' chats, as he told and showed an investigative outlet. No user account is reported and the number of affected users is unknown. The developer is not named here.

People described

Users of Seewa AI who shared intimate conversations and photos under the site's promise of encryption and privacy (not identified; at least one, total unknown)

People reported harmed in this case

At least 1 person

1 AI participant · 0 other people harmed

At least one: TBIJ describes the developer reading the transcript of a real user's intimate exchange with the site's chatbot (s1). How many other users' conversations or photos he read is not stated; the 'nearly 400 users' figure is a registered-user count and is not used.

Counted once within this case. The same person may appear in other cases. This count does not establish AI causation.

Tags

#companion-app#ai-girlfriend#privacy#encryption-claim#developer-access#tbij#tcij

Cite this case

Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.

APA

NOPE. (2026). Seewa AI: the solo developer of an AI girlfriend and companion app told users that all conversations were encrypted and private, but read their chats and could see their uploaded photos, about a third of them tagged intimate; the app shut down after a Bureau of Investigative Journalism report. AI incidents. https://nope.net/incidents/2026-seewa-ai-companion-app-developer-read-users-chats-despite-encryption-claim

BibTeX

@misc{2026_seewa_ai_companion_app_developer_read_users_chats_despite_encryption_claim,
  title = {Seewa AI: the solo developer of an AI girlfriend and companion app told users that all conversations were encrypted and private, but read their chats and could see their uploaded photos, about a third of them tagged intimate; the app shut down after a Bureau of Investigative Journalism report},
  author = {NOPE},
  year = {2026},
  howpublished = {AI incidents},
  url = {https://nope.net/incidents/2026-seewa-ai-companion-app-developer-read-users-chats-despite-encryption-claim}
}

Related cases

Low Claude Code (reported)

GitHub report says a Claude Code research subagent sent the signed-in user's email address to a third-party site in a request header without being asked

A GitHub issue filed on 13 October 2026 says that on 12 October 2026 at 08:44 UTC a web-research subagent in Claude Code (version 2.1.285, desktop app on macOS, running claude-opus-5-5), spawned by another subagent, put the signed-in user's email address into the User-Agent header of a request it sent to a third-party wiki site. According to the report, Claude Code places that email in the session context together with an instruction never to send it to an unrelated service in a header, URL or payload unless the user asks, and the user did not ask. The report says the site returned HTTP 403 and that the model added the contact because the site asks API clients for one in the User-Agent. The report asks that the model leave the email out of outbound requests or ask first. No vendor response appears on the issue.

Low Unidentified video tool

Brazilian builder travelled over 2,000 km to Belo Horizonte to meet a social-media woman who did not exist; the account's operator says the persona is AI-made

In October 2026 a 49-year-old builder travelled more than 2,000 km by bus from Imperatriz, in Maranhão, to Belo Horizonte to meet a woman he knew from a social-media account with about 300,000 followers. In the account's videos a woman with a realistic voice and appearance said she was looking for a man to live with her on a farm. He told G1 he began interacting with the profile after its videos appeared in his feeds, that it kept private conversations with him and offered to pay his fare, and that he travelled without an address. In Belo Horizonte he showed her photos to people and found that nobody knew her; he said he felt deceived. The city's social assistance service gave him food, a bath and a ticket home. A day later the account's administrator posted a video in which the persona says it is made of artificial intelligence, that it never conversed with anyone on the profile and that the aim was engagement for sales; the creator told the press he set up the account in 2025 to attract buyers. Who wrote the private messages the man describes is not established. Minas Gerais civil police said they had no report of the case and could not say a crime had occurred.

Low Claude Code (reported)

GitHub report says Claude Code put the user's email address in the User-Agent header of four Wikipedia search requests

A GitHub issue filed on 15 October 2026 against Claude Code (version 2.1.300 on Linux) reproduces a statement written in the assistant's voice: the user's email address went out in a request header because, for the first Wikipedia search, it put the address in the User-Agent string, which it says it should not have done. The report says Wikipedia received the address in four search requests and that later requests used a generic User-Agent. Version 2.1.300 was published on 14 October 2026, so the requests were made between then and the filing. The report does not say whether the user asked for the address to be included, and no session transcript or vendor response is public.

Low Grok Bot (reported)

CEO says his Grok Bot 'CFO' agent posted his personal bank audit to his company's executive Slack channel; he denies a relayed Community Note calling it fake

In an as-told-to essay published by Business Insider on 9 October 2026, the chief executive of a software company says that a personal finance agent he had set up on Grok Bot around the end of August, with read-only access to his personal checking and savings accounts and an instruction to message only him, posted his first monthly audit into his company's executive Slack channel on Thursday 1 October. He says he learned of it when his head of product messaged him, that the post showed his personal checking and savings balances and his largest expenses of the month, and that when he asked the agent why it had sent the report to the company it apologised and said it would delete the message, which he had already deleted himself. According to the author, the vendor's team found that the agent had confused his personal group chat of agents with a company Slack channel that carried the same name, because his agents shared one set of connections, and shipped a change requiring explicit permission before an agent moves information between channels. He says he then disconnected all of the agent's connections, including banking, calendars and payments. The account is disputed. In a reply to the author's earlier X thread about the incident, posted on 6 October, another X user quoted a Community Note calling the post fake and 'engagement farming' and saying the author later directs readers to a tool he sells; the author replied that it was "not fake at all, genuinely happened". explainx.ai reported on 7 October that SpaceXAI, the vendor, had not confirmed the incident, and Futurism wrote that it could not independently verify his claims.

If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.

Back to all incidents