Anthropic faced fresh scrutiny after publicly shared Claude conversations were exposed in Google search results, while CEO Dario Amodei issued a July 27 open letter defending the company’s position on open-weights models, according to Jiemian News. The incident stemmed from a configuration issue involving robots directives and noindex behavior, and the searchable links surfaced private chats that included a crypto wallet with a private key, legal notes, resumes, possible Social Security numbers, clinical trial summaries and employee details. By July 26, Google had removed the results, while Bing and Brave still displayed them; Anthropic had not publicly commented by July 28.
Jiemian News said the leak added to a series of security and policy controversies surrounding Anthropic. The article said this was the fourth similar incident in a year among closed-model providers, with OpenAI and xAI each also facing one case. It also cited previous scrutiny over Anthropic’s data practices, including a court-approved $1.5 billion settlement in a copyright case, a platform audit showing its Reddit crawler continued visiting after being blacklisted, and a reported removal of hidden tracking code from Claude Code in July.
Anthropic Faces Share-Link Leak and Renewed Open-Weights Backlash
2026-07-28 13:22:37
Disclaimer:
1. The information provided does not constitute investment advice. Investors should make independent decisions and bear all risks themselves.
2. The copyright of this content belongs to the original author. The views expressed herein are solely those of the author and do not represent the stance or position of this website.
Previous article：
Birdai Labs完成400万美元融资Next article：
Morgan Stanley Launches Ethereum and SOL Trusts
