Anthropic officially announced the expansion of its flagship program Project Glasswing.
The market holds its breath. Valuation close to 1 billion USD!
This news hit the light of day just one day after the world was electrified by the confidential filing of the company's documents for its first public offering.
This move follows a spectacular funding round totaling as much as 65 billion USD, which catapulted the AI giant’s valuation to almost 1 billion USD.
For Wall Street this is a clear signal that advanced AI is no longer just a text and meme generator, but becoming the foundation of national security and the hottest commodity on the trading floor.
At the heart of the Project Glasswing initiative beats the digital heart of the Claude Mythos model. Anthropic, without reservations, calls it its most powerful and predatory creation in history. Why? Because it possesses unprecedented, even terrifying, abilities to detect thousands of zero‑day vulnerabilities in just a few weeks.
The model can link seemingly unrelated bugs and create lethal attack vectors from them, something only the most elite and talented hackers could do until now.
In early April the first 50 partners, among the governments of major powers, gained early access to the demo version Mythos Preview to scan their source code.
The results shocked the banking sector; the model easily found a 27‑year‑old vulnerability in the OpenBSD system and hundreds of bugs in the Firefox browser and Linux kernel.
Now the program is accelerating dramatically, opening doors for about 150 new organizations. The new cohort includes strategic sectors that have so far been underrepresented, namely energy, water resources, healthcare, telecommunications, and key computer hardware manufacturers.
A successful attack on these resources would have catastrophic effects on the global economy. According to Anthropic estimates, a strike on a single partner’s infrastructure could directly affect more than 100 million people, bringing paralyzing consequences for international security.
See also: CD Projekt shares before the new ATH? Expert says when to expect DLC for The Witcher 3 and the release of The Witcher 4. “I remain cautious.”
Global coalition and digital arms race
The elite trusted list includes entities from countries closely cooperating with the United States, such as Australia, Canada, France, Germany, Italy, Switzerland, Netherlands, Spain, Belgium, Sweden, India, Japan, New Zealand, and South Korea.
Media leaks (including Financial Times reports) reveal specific brands joining the project, such as Okta, Samsung, SK Telecom, and even NATO’s headquarters in Brussels.
Why is Anthropic under such immense time pressure? Because the competition doesn’t sleep. The company openly admits that within the next 6–12 months other labs will develop commercial or open‑source models with similar, destructive capabilities to Claude Mythos.
OpenAI has already thrown down the gauntlet, releasing its own cybersecurity‑dedicated model GPT‑5.5 Cyber, which is set to undergo extensive testing with selected partners.
See also: The 2026 market is preparing a surprise? “Investors should ensure their portfolios are ready for the upward trend.”
Source: TechCrunch.
