Intercontinental Exchange Joins Anthropic’s Project Glasswing as Cybersecurity Becomes Critical Infrastructure
Intercontinental Exchange joins Anthropic’s Project Glasswing, deploying Claude Mythos Preview to strengthen cybersecurity across critical financial infrastructure.
Intercontinental Exchange (NYSE: ICE), the operator of the New York Stock Exchange and a global portfolio of exchanges, clearinghouses, fixed income data businesses, and mortgage technology platforms, has joined Anthropic’s Project Glasswing. The cybersecurity initiative gives select organizations access to Claude Mythos Preview, an unreleased AI model designed to identify software vulnerabilities before they can be exploited.
The deployment spans ICE’s critical infrastructure, including NYSE operations, clearinghouses, fixed income and data services, and mortgage technology businesses. Unlike many AI announcements centered on productivity gains, this one focuses on defense, resilience, and operational security. Anthropic, the San Francisco-based AI company founded by Dario Amodei and Daniela Amodei, created Project Glasswing to help operators of critical infrastructure strengthen cyber defenses before increasingly capable AI systems become widely available.
The timing matters. Financial markets increasingly run on software, and software increasingly represents the attack surface. As AI systems become more capable of discovering vulnerabilities, organizations that operate critical infrastructure are racing to ensure defenders gain access to those capabilities before attackers do.
What Happened
Intercontinental Exchange announced it has joined Anthropic’s Project Glasswing and deployed Claude Mythos Preview across its critical financial infrastructure. The announcement places ICE among a growing group of organizations receiving controlled access to one of the most advanced cybersecurity-focused AI systems currently available.
According to Anthropic, Project Glasswing was created to help organizations identify and remediate software vulnerabilities before malicious actors can exploit them. Participation is restricted, and access is limited to vetted organizations operating systems with broad societal or economic importance. For ICE, the deployment reaches far beyond a single business unit. The company operates infrastructure that supports trading, clearing, market data distribution, fixed income analytics, and mortgage technology workflows globally. Those systems help form part of the digital backbone of modern financial markets.
That scope transforms the announcement from a technology deployment into an infrastructure story.
About Intercontinental Exchange
Intercontinental Exchange occupies a unique position within global financial markets. Founded by Jeffrey Sprecher in 2000, ICE grew from an electronic energy marketplace into one of the most important operators of financial market infrastructure in the world. The company owns the New York Stock Exchange, operates multiple exchanges and clearinghouses, provides fixed income data and analytics, and serves the mortgage industry through its technology platforms.
Many technology companies talk about ecosystems. ICE operates systems that help power actual economic ecosystems. When markets open each morning, investors see stock prices moving across screens. Behind those screens sits an enormous amount of software responsible for execution, settlement, data distribution, compliance, and risk management. Much of that infrastructure ultimately connects back to organizations like ICE. That reality makes cybersecurity less of an IT function and more of a business continuity requirement.
Why Anthropic’s Project Glasswing Matters
The name Project Glasswing reflects vulnerabilities that remain hidden in plain sight. Anthropic built the initiative around a simple but increasingly urgent observation: advanced AI systems are becoming exceptionally good at finding software flaws. The same capability that can help defenders secure systems can also help attackers discover weaknesses at unprecedented speed.
Rather than broadly releasing Claude Mythos Preview, Anthropic restricted access through Project Glasswing. The model remains tightly controlled because of its cybersecurity capabilities and potential misuse risks. The company reports that participating organizations have already identified more than 10,000 high- and critical-severity vulnerabilities through the program. That figure reveals an uncomfortable truth about modern infrastructure. The challenge is no longer whether vulnerabilities exist. The challenge is finding them before somebody else does.
Leadership Signals Behind the Announcement
The executives quoted in the announcement reveal how ICE views the initiative. Ben Jackson, President of Intercontinental Exchange, emphasized secure, auditable deployment for regulated industries and highlighted the role of AI in improving vulnerability detection across critical systems. Lynn Martin, President of NYSE Group and Chair of ICE Fixed Income & Data Services, focused on protecting the infrastructure supporting global capital markets.
Neither statement centered on innovation theater. Both centered on resilience. That distinction matters because financial infrastructure operators face a different set of incentives than consumer technology companies. Reliability is not a feature. Reliability is the product.
Market Context: The New Cybersecurity Arms Race
Cybersecurity has entered a new phase. For years, organizations worried about ransomware groups, nation-state actors, and supply-chain attacks. Now another variable has entered the equation: increasingly capable AI systems.
The industry conversation often focuses on AI-generated content, coding assistants, and productivity gains. Far less attention is paid to what happens when AI becomes exceptionally effective at finding vulnerabilities hidden inside millions of lines of code. That concern sits at the core of Project Glasswing. Anthropic has argued that comparable capabilities could become more broadly available within the next 6–12 months. If that timeline proves accurate, organizations responsible for critical infrastructure may have a limited window to strengthen defenses before similar capabilities become commonplace.
Project Glasswing has also expanded beyond financial services, bringing additional organizations from sectors including healthcare, communications, utilities, and other critical infrastructure environments into the initiative. The organizations moving first are sending a clear signal: cybersecurity is becoming a strategic infrastructure investment rather than a defensive afterthought.
What This Signals
The deeper story is not about Anthropic, and it is not even primarily about ICE. It is about the growing convergence of AI, cybersecurity, and critical infrastructure. Financial markets, energy systems, healthcare networks, communications platforms, and government systems increasingly depend on software. As those systems become more complex, vulnerability discovery becomes both more important and more difficult.
Project Glasswing reflects a broader shift toward proactive defense. Organizations are no longer waiting for threats to emerge before responding. They are seeking tools capable of identifying weaknesses before adversaries can exploit them. For operators, investors, and technology leaders, that shift may become one of the defining infrastructure themes of the next decade. The firms that learn to integrate advanced AI into cybersecurity operations without sacrificing governance, auditability, or resilience could establish a meaningful advantage long before the broader market recognizes the stakes.
Frequently Asked Questions
What is Project Glasswing?
Project Glasswing is Anthropic’s cybersecurity initiative that provides vetted organizations with access to advanced AI tools for vulnerability discovery and remediation.
Why did Intercontinental Exchange join Project Glasswing?
Intercontinental Exchange joined Project Glasswing to strengthen cybersecurity across exchanges, clearinghouses, data services, and mortgage technology platforms.
What is Claude Mythos Preview?
Claude Mythos Preview is an unreleased Anthropic AI model designed to identify software vulnerabilities and support cybersecurity defense operations.
Why is financial infrastructure a cybersecurity target?
Financial infrastructure supports trading, clearing, market data distribution, and other essential market functions, making it a high-value target for cybercriminals and nation-state actors.
Who founded Anthropic?
Anthropic was founded by Dario Amodei and Daniela Amodei alongside former OpenAI researchers and executives.
What does Intercontinental Exchange own?
Intercontinental Exchange owns the New York Stock Exchange and operates exchanges, clearinghouses, fixed income data services, and mortgage technology platforms.
How does AI improve cybersecurity?
AI can help identify software vulnerabilities, analyze code, prioritize risks, and accelerate remediation efforts across large technology environments.
Why does this announcement matter beyond ICE?
The announcement signals growing adoption of advanced AI cybersecurity tools across organizations responsible for critical economic and operational infrastructure.









