According to a Bloomberg News report dated Tuesday, April 22, 2026, citing documents and people familiar with the matter, Anthropic’s AI model Claude Mythos Preview was accessed by a small batch of unauthorized users on the same day that Anthropic first publicly announced its plan to open up Mythos testing to a limited number of enterprises.
Bloomberg Report Details: A Small Number of Users Received Access to Claude Mythos Preview
According to Bloomberg’s report, in a private online forum, a small number of users received access to Claude Mythos Preview on the same day Anthropic announced its plan to open Mythos testing to a limited number of businesses; since then, that user group has been using Mythos regularly. However, Bloomberg’s report said that the purpose of use was not for cybersecurity purposes. Bloomberg stated that, in the course of reporting on this investigation, the access party provided screenshots and an on-site demonstration of model use.
An Anthropic spokesperson confirmed that an investigation has been initiated and did not provide further details on the investigation’s progress.
Background on Claude Mythos and Project Glasswing
According to Anthropic’s public materials, Claude Mythos Preview was announced on April 7, 2026 as part of Anthropic’s “Project Glasswing.” Project Glasswing is a controlled program that allows certain organizations to use an unreleased Claude Mythos Preview to perform defensive cybersecurity work. Anthropic previously said that Mythos has unprecedented capabilities to identify digital security vulnerabilities and potential misuse, and that it has drawn the attention of regulators.
Frequently Asked Questions
When did Bloomberg report that Claude Mythos Preview was accessed without authorization?
According to a Bloomberg News report dated Tuesday, April 22, 2026, citing documents and people familiar with the matter, Claude Mythos Preview was accessed without authorization by a small number of users in a private forum on the same day Anthropic announced plans to open testing to limited enterprises.
What official response did Anthropic provide to this incident?
According to an Anthropic spokesperson statement cited by Bloomberg, Anthropic said: “We are investigating a report that claims someone accessed Claude Mythos Preview without authorization through an environment provided by one of our third-party vendors.”
Which program does Claude Mythos Preview belong to, and what is it designed for?
According to Anthropic’s public materials, Claude Mythos Preview was announced on April 7, 2026 and belongs to the “Project Glasswing” program, which allows certain organizations to use an unreleased model for defensive cybersecurity work.
Disclaimer: The information on this page may come from third parties and does not represent the views or opinions of Gate. The content displayed on this page is for reference only and does not constitute any financial, investment, or legal advice. Gate does not guarantee the accuracy or completeness of the information and shall not be liable for any losses arising from the use of this information. Virtual asset investments carry high risks and are subject to significant price volatility. You may lose all of your invested principal. Please fully understand the relevant risks and make prudent decisions based on your own financial situation and risk tolerance. For details, please refer to
Disclaimer.
Related Articles
ASMPT Projects Strong Q2 as AI Demand Drives Revenue to $540M–$600M
ASMPT lifts Q2 revenue guidance to $540–600m on AI-driven demand; Q1 revenue was $507.9m with record bookings. Growth from TCB and HBM4, photonics, wire/die bonding, and AI-driven SMT demand; SMT segment options under review.
ASMPT announced a higher second-quarter revenue outlook of $540 million to $600 million, driven by AI-related semiconductor demand, following first-quarter revenue of $507.9 million and record bookings. Growth is led by Thermo-Compression Bonding tools for logic chips, HBM4 modules, and photonics for data centers, with contributions from wire/die bonding and AI-driven SMT demand. The company reported no material impact from the Iran conflict and is evaluating strategic options for its SMT Solutions Segment to focus on the higher-growth Semiconductor Solutions Segment.
GateNews12m ago
Claude Mythos finds 271 vulnerabilities in Firefox, defenders may have a decisive advantage
Mozilla announced on Tuesday that an early version of Anthropic’s Claude Mythos AI model, identified 271 security vulnerabilities in the Firefox browser during internal testing, and that all of the issues have been fixed this week. While Mozilla said it was shocked, it also noted that the results suggest a fundamental shift may be underway in the field of online security, and that defenders may be about to narrow the long-standing advantage attackers have enjoyed.
MarketWhisper17m ago
SpaceX IPO Warning: Elon Musk’s Space AI Data Center Business Viability Is in Doubt
According to a Reuters report on April 22, in a draft of its S-1 registration statement to be filed soon, SpaceX issued a warning to potential investors, saying that its plans to develop space-based AI data centers and to industrialize the Moon and interplanetary space are “still in an early stage, involve significant technical complexity and unproven technology, and may not achieve commercial viability,” sharply contrasting with the optimistic attitude its CEO, Musk, has publicly expressed this year.
MarketWhisper46m ago
Claude Code reportedly suspends the Pro plan rollout and says you need to subscribe to Max to use it. An Anthropic executive says it’s still being tested.
Recently, multiple users have found that the Anthropic website appears to have removed Claude Code from Pro (20 USD/month) and instead made it available under Max (starting at 100 USD/month). The official statement says this is only a test aimed at about 2% of new prosumer registrations, and existing Pro and Max users are not affected. This move is seen as a re-evaluation of the cost and pricing strategy for high-consumption features; in the future, such features may be separated from mainstream plans and move toward higher-tier, tiered pricing.
ChainNewsAbmedia1h ago
ChatGPT Images 2.0 is here! Text generation accuracy is greatly improved, making it easy to create marketing posters
OpenAI releases ChatGPT Images 2.0. Text generation is more accurate, and posters and portrait designs look more beautiful. A new “thinking mode” enables real-time web browsing, batch output, and self-review. With a single prompt, you can generate up to eight images; it can maintain character and style consistency. The maximum resolution is 2K, and the aspect ratio can be selected from 3:1 to 1:3. Asian-language optimization is significantly improved. It offers an API and paid commercial plans. Knowledge cutoff is December 2025, and even complex compositions still require a few minutes.
ChainNewsAbmedia1h ago
OpenAI Commits Up to $1.5B Investment in New Private Equity Joint Venture
Gate News message, April 22 — OpenAI has committed to investing up to $1.5 billion in a new joint venture with a private equity firm, aiming to compete with Anthropic and capture the growing market for enterprise artificial intelligence tools. The company will initially invest $500 million in equity
GateNews1h ago
