Fleet Uses AI to Close Patching Gap

Fleet, a developer of open-source device management software, has joined Project Glasswing, a restricted access program run by Anthropic. This partnership gives the company early access to Claude Mythos Preview, a frontier AI model that has not yet been released to the general public.

The primary goal of this collaboration is to determine how advanced artificial intelligence can protect the infrastructure that manages company devices. Specifically, Fleet aims to narrow the dangerous gap between the speed at which cyberattacks emerge and the time IT teams currently require to deploy defensive patches and updates.

Bridging the Critical Speed Gap

According to research published by Fleet, a significant majority of organizations take more than a day to roll out critical security fixes. In contrast, malicious actors often exploit newly discovered vulnerabilities within just a few hours. This disparity creates a window of vulnerability where networks are exposed. By integrating AI capabilities, Fleet hopes to automate and accelerate the identification and remediation of these risks, allowing security teams to respond in near real-time rather than waiting for manual processes to complete.

As reported by 9to5Mac, this initiative is not just about internal efficiency. It is a direct response to the changing landscape of cyber threats, where the speed of attack has outpaced traditional defensive strategies. The company believes that leveraging AI can fundamentally alter the balance of power between defenders and attackers.

Open Source Transparency Commitments

A key differentiator in this project is Fleet’s commitment to transparency. Because the core of its product is open source, the company states that it will share its findings and learned best practices with the broader security community. This approach ensures that the benefits of the AI integration are not limited to Fleet’s direct customers but are available for anyone building or maintaining device management systems.

Mike McNeil, the CEO and co-founder of Fleet, emphasized that this move aligns with the company’s core values of community and openness. He noted that celebrating the company’s founding anniversary with a renewed focus on product security feels particularly fitting. The goal is to provide practical insights into where frontier AI can meaningfully improve security workflows for everyone.

Testing Real World Workflows

The research will focus on specific operational challenges, such as understanding the current state of endpoints and prioritizing which issues require immediate remediation. By testing these AI capabilities against real-world workflows before general release, Fleet aims to ensure that the technology is both safe and effective. This controlled environment allows for rigorous evaluation of how AI can handle complex security operations without the risks associated with a public launch.

Industry partners have expressed support for this direction. Gabe Chan, a Staff IT Engineer at Abridge, noted that the instantaneous nature of modern AI-driven attacks has changed expectations for technology providers. He highlighted the importance of vendors continuously improving their ability to identify and respond to risk, praising Fleet for investing in research that strengthens its own security while contributing valuable knowledge back to the industry.