The Essential Complementarity of VALIDY TECHNOLOGY with CLAUDE MYTHOS by ANTHROPIC within the GLASSWING Project Framework
This project consists of a consortium bringing together major players in cloud computing, cybersecurity, software, and finance (launch partners of the project).
• cloud & infrastructure (AWS, Microsoft, Google)
• manufacturers & hardware (Apple, NVIDIA, Broadcom)
• cybersecurity (CrowdStrike, Palo Alto Networks, Cisco)
• critical finance (JPMorgan Chase)
• open source (Linux Foundation)
The idea is to secure the core of global digital infrastructure before equivalent capabilities are exploited offensively.
From probabilistic cybersecurity to deterministic cybersecurity
For more than thirty years, global cybersecurity has been primarily based on a probabilistic approach to risk. Protection systems aim to reduce the likelihood of compromise through the accumulation of defensive mechanisms: code audits, software patches, behavioral detection, antivirus, heuristic analysis, or artificial intelligence applied to vulnerability discovery.
Recent initiatives (April 2026) in automated code analysis using AI, such as the CLAUDE MYTHOS project developed by ANTHROPIC, represent the culmination of this logic. They make it possible to identify potential flaws more quickly, including unknown vulnerabilities (zero-day), and significantly improve overall software quality.
However, this approach remains intrinsically probabilistic: it reduces risk without being able to eliminate it. No analysis system, whether human or algorithmic, can demonstrate the absolute absence of vulnerability in complex software. Moreover, even a perfectly audited program can be altered after deployment through compromise of the execution chain.
Why would the global deployment by ANTHROPIC of the new CLAUDE MYTHOS AI version be so dangerous today?
An AI model as powerful as CLAUDE MYTHOS, if it fell into the wrong hands, would provide offensive capabilities to states, criminal groups, or malicious hackers far exceeding current defensive capabilities.
And we know that most essential services are digitized. Therefore, the dissemination of CLAUDE MYTHOS would democratize hacking on an unprecedented scale, effectively giving cyber attackers the equivalent of a nuclear weapon, while cyber defenders would be left with insufficient—if not negligible—means.
And if ANTHROPIC has managed to build such a powerful AI with such cyber capabilities, other companies and nations will undoubtedly achieve the same very soon.
And they will not all be Western…
This is why the complementarity of VALIDY TECHNOLOGY becomes as necessary as it is indispensable within the GLASSWING consortium.
VALIDY TECHNOLOGY introduces a paradigm shift by moving the point of trust from software to its hardware execution. As a Runtime Root of Trust embedded within the microprocessor, the technology does not attempt to anticipate attacks but to physically prevent any unauthorized modification of software behavior from producing a real effect.
The continuous verification of execution integrity, performed in real time through elementary operations with extremely low computational cost, transforms cybersecurity: the question is no longer merely “is an attack likely?” but “can an attack still produce an outcome?”
This transition marks the shift:
• from cybersecurity based on threat detection,
• to cybersecurity based on the non-alterability of execution.
In this model, predictive code analysis and hardware-based execution guarantees become complementary. Artificial intelligence reduces the introduction of vulnerabilities; the Runtime Root of Trust prevents their operational exploitation.
For states, this evolution goes beyond the technical dimension. It opens the way to a stabilization of cyberspace comparable to that historically provided by certain critical physical infrastructures: an environment where offensive capability remains, but where massive and invisible system compromise becomes structurally more difficult.
Thus emerges a deterministic cybersecurity, in which trust no longer relies solely on analysis or monitoring, but on the intrinsic properties of computation execution itself.
CLAUDE MYTHOS by ANTHROPIC and VALIDY TECHNOLOGY: toward a two-layer cyber architecture (AI and Runtime Root of Trust)
This dynamic reflects a collective realization of a now established fact: human code analysis is no longer sufficient given the complexity and scale of modern digital systems. CLAUDE MYTHOS by ANTHROPIC provides a response adapted to this scale by automating vulnerability detection, including the most advanced ones, and significantly reducing the probability of introducing exploitable flaws.
However, this first layer, as essential as it is, does not resolve a structural limitation of cyberspace: the impossibility of guaranteeing that software will remain intact after analysis, deployment, and throughout its execution. The history of sophisticated attacks, particularly those targeting supply chains or execution environments, demonstrates that compromise can occur at any point in the digital lifecycle.
This is precisely where VALIDY TECHNOLOGY intervenes.
As a Runtime Root of Trust embedded at the core of microprocessors, VALIDY TECHNOLOGY constitutes a second layer of security, no longer based on analysis or detection, but on guaranteeing execution integrity itself. It ensures, in real time, that software behavior strictly matches what was validated at compilation, making any operational alteration immediately detectable and neutralizable.
The combination of these two approaches creates a coherent and comprehensive architecture:
• a first layer, cognitive and predictive, based on artificial intelligence, aimed at identifying and reducing vulnerabilities upstream;
• a second layer, hardware-based and deterministic, ensuring that these vulnerabilities cannot produce effects during execution, even if they persist or are introduced later.
This two-layer architecture represents a paradigm shift. It no longer merely seeks to reduce risk but to structurally control its consequences.
For states, this complementarity offers major strategic value. It enables the development of cybersecurity that is both proactive and resilient, capable of adapting to highly hostile environments while maintaining sovereign control over critical infrastructures. It also paves the way for the progressive standardization of a trust chain covering the entire software lifecycle, from design to actual execution.
From this perspective, VALIDY TECHNOLOGY should not be seen as competing with initiatives such as CLAUDE MYTHOS by ANTHROPIC, but as their natural extension in synergy at the most fundamental level of computing. Together, they outline the contours of a next-generation cybersecurity infrastructure, adapted to contemporary challenges and to the strategic balances of cyberspace.
Toward an implicit Western cyber doctrine
The simultaneous emergence of large-scale AI-driven code analysis capabilities, through initiatives such as the GLASSWING project, along with the development of execution integrity technologies like VALIDY TECHNOLOGY, does not reflect isolated developments. Rather, it signals the gradual emergence of an implicit cyber doctrine among Western technological powers.
This doctrine, still informal, is based on a coherent articulation between two complementary principles. On one hand, the use of artificial intelligence to anticipate, detect, and reduce vulnerabilities at scale. On the other, the establishment of execution trust anchored in hardware, ensuring that digital systems behave as validated, regardless of operating conditions.
This dual movement responds to a strategic necessity: in the face of the industrialization of attacks and the growing complexity of systems, no single approach can ensure overall security. The combination of predictive cybersecurity and deterministic cybersecurity thus constitutes a systemic response adapted to contemporary environments.
Beyond the technical dimension, this convergence outlines the contours of a common framework between states, industrial players, and operators of critical infrastructures. It fosters the emergence of de facto standards, based not on formal agreements but on the gradual adoption of shared practices and architectures.
In this perspective, the association of technologies such as CLAUDE MYTHOS and VALIDY TECHNOLOGY could help structure a transnational trust infrastructure, compatible with sovereignty requirements while facilitating interoperability among allies. It thus opens the way to a relative stabilization of cyberspace, in which control over digital systems depends as much on their design as on the technical impossibility of altering them during execution.
Without constituting a formal doctrine, this dynamic already resembles a form of strategic convergence, destined to play a structuring role in the international digital balance of the coming decades.