BeyondTrust has joined Anthropic's Project Glasswing, placing the cybersecurity company in a cohort focused on securing software used in critical infrastructure.
Project Glasswing is a collaborative initiative focused on software codebases that underpin systems used across government, healthcare, commerce, and essential services. BeyondTrust's inclusion reflects the role of its products in environments where privileged access controls are closely tied to operational resilience.
Software security
Through the programme, BeyondTrust will use Claude Mythos Preview, an AI model, to identify, validate, and remediate software vulnerabilities across its products. That work will also cover the BeyondTrust Pathfinder Platform, the company's architecture for securing privilege across human, machine, and agentic identities in endpoint, cloud, and on-premise environments.
The move underscores growing attention to identity and privilege management as a cybersecurity issue, particularly as organisations add more automated systems and AI agents to core operations. Security teams have long viewed privileged accounts as a primary target for attackers because those credentials can provide broad access to systems, data, and administrative tools.
BeyondTrust argued that the growth in machine identities and AI-driven processes is adding complexity to that challenge. It said AI agents are becoming a rapidly expanding class of privileged identities, often with access to sensitive business processes and systems while remaining less governed than many human user accounts.
Vulnerability work
Anthropic said Claude Mythos Preview has already helped programme participants surface more than 10,000 high- and critical-severity vulnerabilities. BeyondTrust did not disclose how many issues it expects to find in its own software or whether the work will be limited to internally developed code.
According to the company, BeyondTrust serves more than 20,000 customers, including 75 of the Fortune 100. Its products are used to manage and secure privileged access, a category that has gained prominence as companies seek tighter controls over administrators, service accounts, workloads, and newer forms of software-based identity.
Janine Seebeck, Chief Executive Officer, BeyondTrust, framed the company's participation as part of that broader shift.
"Millions of the world's most important workloads are protected by privilege controls that we build and maintain. That is a responsibility we take seriously," said Janine Seebeck, Chief Executive Officer, BeyondTrust.
"As AI reshapes both software development and cyber defence, the organisations responsible for securing critical infrastructure must continuously raise the bar. Project Glasswing gives us an opportunity to do exactly that by continuing to strengthen the security and resilience of the software our customers rely on to protect all human, machine, and agentic identities," said Seebeck.
Identity focus
The announcement comes as cybersecurity vendors and infrastructure operators face pressure to tighten development practices in response to faster attack cycles. BeyondTrust said the time between the discovery of a vulnerability and its weaponisation has contracted sharply as AI tools are used on both sides of the security equation.
That trend has increased interest in using AI models not only for coding assistance but also for defensive security work such as code analysis, flaw discovery, and remediation support. For suppliers of software embedded in critical environments, the ability to detect vulnerabilities earlier has become a strategic concern as customers and regulators pay closer attention to software supply chain risk.
BeyondTrust's participation also shows how identity security is being linked more directly to infrastructure protection. Rather than focusing only on network perimeters, vendors in this area increasingly argue that access rights, administrative privileges, and non-human identities represent a central line of defence.
Shared response
Marc Maiffret, Chief Technology Officer, BeyondTrust, said the threat environment requires shared action across the industry.
"The threats ahead are bigger than any one vendor, and the response has to be shared," said Marc Maiffret, Chief Technology Officer, BeyondTrust.
"We are honoured to stand with Anthropic and the other members of Project Glasswing, applying Mythos to our own code to further strengthen the security of the products our customers depend on, and doing our part in a defence no one can mount alone," said Maiffret.