Insider threats have historically been viewed through a fairly straightforward lens: a person with legitimate access misuses it, intentionally or otherwise. But as organizations rely on more connected systems and autonomous technologies, the scope of insider risk is changing.
An employee with access to a restricted facility and an AI agent operating inside a software development environment may seem like separate security concerns. Both, however, represent trusted access that needs to be managed and monitored. As Insider Threat Awareness Month comes to a close, organizations should consider how insider risk now stretches across physical spaces and software environments.
Physical security starts with knowing who has access
In physical environments, managing insider risk often begins with basic questions around access. Organizations need to know who can enter sensitive spaces, whether those permissions still align with an individual’s role and whether security teams have the information needed to investigate suspicious activity.
Chris Robertson, Director, Technical Consultant Team, Luminys:
“After more than 25 years working in physical security, one thing I’ve learned is that managing insider risk often comes back to the fundamentals. Who has access to a physical space? Are those access permissions still appropriate for that person’s role? And can the security team quickly verify what happened when an incident occurs? Effective access control, video security and strong operating procedures give organizations the visibility they need to identify potential insider threats, investigate incidents and act on what they can verify.
Technology has improved considerably, especially when video security, access control and search tools work together, but good security still depends on how teams design and manage the system every day. Teams must review permissions. Video needs to be usable when an investigation happens. Operators need tools they understand and trust. Insider Threat Awareness Month is an essential moment to take a closer look at those basics and make sure the physical security program is doing what the organization expects it to do.”
AI agents are becoming part of the access equation
Trusted access is also changing within software environments. AI agents are increasingly able to interact directly with code, data and development infrastructure, giving organizations a new category of identity to consider when assessing insider risk.
Unlike traditional software tools that execute narrowly defined tasks, autonomous agents can make decisions and take actions across multiple systems. That makes understanding what they can access, what they can change and what enters the software supply chain through their activity increasingly important.
Hari Srinivasan, Vice President of Products & Strategy, Lineaje:
“Insider Threat Awareness Month can no longer focus solely on human actors. Today, enterprise insider risk extends to autonomous software agents operating inside development pipelines. Organizations are granting AI agents privileged access to private repositories, CI/CD systems, package managers, skills repositories, and systems of records. Functionally, these agents act as credentialed insiders: they select third-party packages, write and refactor code, and push commits directly into trusted environments. Further, they can access, process, transmit, and transform data from systems of record, as well as from external models and other agents.
When an autonomous agent is untracked or misgoverned, it not only introduces bugs, but also injects supply chain risk directly into production builds at machine speed, and could leak sensitive and business critical data. To contain this emerging vector, security teams must treat every AI agent as an insider and a potential actor involved in the software supply chain. That demands complete visibility into package, model, skills, MCPs and third-party agent provenance, along with continuous assessment of agent-generated code. Enforce automated policy control and governance that neutralize vulnerable or malicious components and actions before impact.”
Different environments, similar questions
The risks associated with a secured facility and a software development pipeline are different, but they raise similar questions for security teams. Who or what has access? Does that access remain appropriate? What actions can a trusted identity take? And does the organization have enough visibility to determine what happened if something goes wrong?
For physical security teams, answering those questions can involve reviewing permissions and connecting access activity with video. For software security teams, it can mean tracking the systems and components AI agents interact with and applying governance to their actions.
Expanding the insider threat playbook
The close of Insider Threat Awareness Month is a timely point for organizations to assess whether their insider risk programs reflect how access is changing. Employees and contractors remain part of that equation, but they are no longer the only trusted actors operating across critical environments.
As organizations introduce new systems and grant greater autonomy to AI agents, insider risk programs will need to keep pace. Maintaining clear permissions, visibility and governance across both physical and digital access can help security teams understand where trust has been granted and whether the controls around it are still doing their job.
