Anthropic is moving its AI security work into higher-stakes territory. The company is expanding Project Glasswing, its security vulnerability program, while opening broader access to Mythos for 150 organizations across 15 countries. The focus is clear: protect critical infrastructure before a cyberattack turns into a public emergency.
The expansion targets sectors where downtime is not just inconvenient. Power grids, water systems, hospitals, and communications networks are all on the list. Anthropic says the organizations involved support services that could affect as many as 100 million people if disrupted by a major cyber incident.
Anthropic Project Glasswing Expands to Critical Infrastructure Cybersecurity
Project Glasswing is designed to help organizations identify and address security vulnerabilities before attackers exploit them. By scaling the program internationally, Anthropic is placing its AI security tools closer to the environments where cyber risks carry real-world consequences.
Critical infrastructure has become one of the most attractive targets for ransomware groups, state-backed hackers, and criminal networks. A compromised utility or hospital system can create cascading damage, from delayed medical care to outages that ripple across entire regions. That makes vulnerability discovery, response planning, and defensive testing more urgent than ever.
Claude Mythos Access Reaches 150 Organizations in 15 Countries
The wider rollout of Mythos gives selected organizations access to Anthropic’s AI-supported security capabilities. While the company has not positioned the expansion as a consumer product launch, the move signals how AI developers are increasingly working with governments, infrastructure operators, and enterprise security teams.
For organizations managing essential services, AI can help sift through large amounts of security data, surface unusual patterns, and speed up investigation workflows. The value is not just in finding a weakness. It is in reducing the time between detection and action, especially when a service outage could affect millions of people.
Why AI Security Tools Matter for Power, Water, Healthcare, and Communications
The sectors included in Anthropic’s expansion share a common problem: they rely on complex, interconnected systems that were not always built with modern cyber threats in mind. Many infrastructure operators must defend legacy technology, cloud services, connected sensors, remote access tools, and third-party software all at once.
That creates a wide attack surface. A single neglected vulnerability can become the entry point for a much larger breach. AI security systems like those Anthropic is developing may help teams prioritize the most dangerous weaknesses, simulate attack paths, and improve incident readiness without forcing analysts to manually inspect every signal.
Anthropic’s Critical Infrastructure Push Reflects a Bigger AI Cybersecurity Shift
Anthropic’s announcement arrives as AI companies face growing pressure to prove that advanced models can be used safely in sensitive environments. Cybersecurity is one of the areas where the promise and risk of AI sit side by side. The same technology that helps defenders spot flaws can also be misused by attackers if safeguards are weak.
That is why controlled access programs matter. By working with a defined group of vetted organizations, Anthropic can gather feedback from real infrastructure environments while keeping deployment more focused than an open public release.
What This Means for Global Cyber Resilience
The expansion of Project Glasswing and Mythos is not a silver bullet. No AI platform can replace strong security teams, disciplined patching, network segmentation, and tested response plans. But it could give critical infrastructure operators another layer of support at a time when attacks are becoming more frequent and more disruptive.
For Anthropic, this is also a strategic step beyond general-purpose AI. The company is showing that its tools are not only meant for productivity and research, but also for defending the systems people depend on every day. If the program delivers measurable improvements, it could become an important model for how AI cybersecurity partnerships are built across borders.
Tags: #Anthropic #ClaudeMythos #ProjectGlasswing #Cybersecurity #CriticalInfrastructure