June 03, 2026 – The AI firm expands Project Glasswing to 150 organizations. It targets critical infrastructure sectors that serve over 100 million people.
In Summary
Anthropic is scaling Claude Mythos to 150 organizations across 16 countries globally
The expansion targets power, water, healthcare, and communications infrastructure sectors
A cyberattack on these partners could potentially affect over 100 million people worldwide
The move follows Anthropic’s confidential IPO filing at a nearly $1 trillion valuation
OpenAI has launched a rival cybersecurity model called GPT-5.5-Cyber for testing
Anthropic announced a major expansion of Project Glasswing on Tuesday. The initiative now reaches 150 organizations across 16 countries. These partners will gain access to Claude Mythos, the firm’s most powerful AI model.
The expansion follows a pivotal moment for the company. Anthropic filed confidentially for an initial public offering on Monday. The firm recently raised $65 billion at a valuation nearing $1 trillion. This financial backing clearly supports its ambitious infrastructure security goals.
Project Glasswing first launched in early April 2026. At that time, Anthropic gave 50 initial partners access to Claude Mythos. The U.S. government was among these early adopters. The model can identify thousands of zero-day vulnerabilities within weeks. This capability makes it a powerful tool for national security teams.
However, the initial cohort lacked representation from key sectors. Power, water, healthcare, and communications were notably absent. Furthermore, many critical codebases remained unprotected. The latest expansion directly addresses these gaps with targeted coverage.
The expanded list now includes vital industries. Additionally, it covers nonprofits that maintain widely used codebases. A successful attack on these systems could prove catastrophic. Anthropic estimates that major breaches could affect over 100 million people. Therefore, the stakes for this program are incredibly high.
The geographic scope spans three major regions. European partners include Germany, France, Italy, and Spain. The Netherlands, Belgium, Sweden, and Switzerland also joined. Meanwhile, the Asia-Pacific region includes Japan, South Korea, India, Australia, and New Zealand. North America adds Canada alongside existing U.S. partnerships.
Key Partners Join the Initiative
Several high-profile organizations have confirmed participation. Okta, Samsung, SK Hynix, and SK Telecom are among the confirmed partners. Furthermore, NATO and the EU’s ENISA agency have signed up. These additions lend significant credibility to the overall initiative.
Many partners maintain codebases that other organizations rely upon. A successful attack on their systems could cascade globally. Anthropic wants to prevent such scenarios through proactive AI scanning. The model identifies vulnerabilities before malicious actors can exploit them.
Competition Heats Up in AI Cybersecurity
Meanwhile, the competitive landscape is heating up quickly. OpenAI recently released its own cybersecurity model called GPT-5.5-Cyber. The company has rolled it out to numerous partners for testing. This move signals a broader industry trend toward AI-powered security tools.
Anthropic expects rivals to match Mythos capabilities soon. The company is therefore racing to establish strong safeguards. Project Glasswing serves as its primary vehicle for this effort. The firm wants to set industry standards before competitors catch up.
Strategic Timing Ahead of Public Debut
The timing of this expansion is strategically significant. Anthropic filed for its IPO just one day prior. The cybersecurity push could strengthen its position with potential investors. Furthermore, it demonstrates real-world applications for Claude Mythos. This tangible utility may support a strong public market debut.
Critical infrastructure remains a top global priority today. Cyberattacks on power grids and water systems have increased recently. Healthcare networks also face growing threats from ransomware groups. Therefore, AI-powered defense tools have become absolutely essential. Governments worldwide are seeking stronger protection for these vital systems.
The 16-country coalition represents a unified international response. Partners span multiple continents and diverse economic systems. This diversity suggests broad global confidence in Anthropic’s technology. Additionally, it creates a powerful network effect for threat detection. Shared intelligence across borders can prevent larger cascading attacks.
Looking Ahead
Looking ahead, Anthropic plans further rapid expansion. The company aims to scale well beyond the current 150 partners. More countries will likely join the initiative throughout 2026. However, regulatory approval may vary significantly by jurisdiction. Data sovereignty rules could limit some international partnerships.
Nonetheless, the trajectory appears clear for industry observers. AI is becoming central to global cybersecurity defense efforts. Anthropic has positioned itself at the forefront of this market shift. The success of Project Glasswing will determine its long-term market position.
