The rapid adoption of AI coding assistants has fundamentally altered the landscape that chief information security officers navigate. Engineering teams across enterprise IT are now writing their own software with the help of generative AI, spinning up autonomous agents that act on their behalf and assigning those agents the same access privileges their human creators hold. This shift has pulled the role of the CISO into territory that did not exist two years ago. Speaking at a recent cybersecurity conference, Hrvoje Englman, CISO at Span, a company that provides IT services to enterprise clients, described how this innovation is reshaping what defenders worry about most.
Span’s workforce includes a sizable population of developers alongside a larger group of engineers, and the engineers have become the new variable. With AI-assisted coding, they are building applications and personal agents to automate parts of their own jobs. Each new agent inherits the identity of its creator, and those identities are typically over-provisioned. Least privilege remains an aspiration that is difficult to enforce in production environments. Instead of blocking this progress, Englman emphasizes enablement. “I cannot be the blocker,” he said. “You cannot block progress. People will find ways around it.” His priority is enabling secure use of AI inside the company rather than prohibiting it.
The Bus-Factor Problem Multiplies
The risk extends beyond simple access control. When a single engineer automates a business process using five interacting agents and then leaves for another job, the organization inherits an undocumented system that nobody understands. Englman called this an inversion of the traditional bus-factor problem. Previously, a key person leaving created a knowledge gap. Now the agents they built keep running, and the company has no record of what they do or why. This undocumented state increases the attack surface and makes incident response significantly harder, as defenders cannot trace the logic or dependencies of the agents.
The bus-factor problem is compounded by the speed at which these agents are deployed. Developers and engineers often bypass formal review processes, believing that AI-generated code is inherently safe because it follows best practices. Yet without proper oversight, each new agent can introduce subtle vulnerabilities, such as unintended data exposure or privilege escalation paths. The security team must then play catch-up, mapping out agent behaviors after they are already in production. Englman noted that this creates a scenario where the company is effectively running blind in parts of its own infrastructure.
Defender’s Leverage Is Real, with Limits
AI has also produced concrete gains in defensive work. Englman pointed to log analysis as one area where the value is immediate. Feeding hundreds of megabytes of log files into an AI tool and asking it to surface anomalies or pivot on an IP address compresses work that previously took analysts hours. Policy drafting is another use case. Generating a first draft from internal context can cut a three-day task to a single day, and the time savings compound across a workforce. These productivity boosts allow security teams to focus on higher-level analysis and strategic planning rather than rote tasks.
However, he drew a sharp line on the vendor pitch for autonomous AI-driven security operations centers. The idea of defensive AI battling offensive AI in real-time, with no humans in the loop, does not match what is achievable now. Log ingestion remains the hardest part of running a SOC, and detection engineering still depends on people who can explain why an alert fired. “You get an alert, but your analyst doesn’t understand the alert,” Englman said, describing the failure mode he sees in teams that lean too heavily on automated tooling. “And you have two million alerts, and then what?” Autonomous isolation of systems remains out of reach because the AI does not understand the business process. Decisions about when to shut down a critical service get escalated to senior leadership during real incidents, and that judgment stays with humans.
He also pushed back on the industry framing of breaches. Most of the largest incidents trace back to phishing and credential theft. Vendors selling AI-powered SOCs as a defense against nation-state actors are addressing a smaller part of the problem than their marketing suggests. The fundamental challenge remains basic hygiene: patching vulnerabilities, enforcing multi-factor authentication, and training users to recognize social engineering attempts.
The Threat Model for a Services Provider
Span sells IT services to enterprise clients, which doubles its exposure. The company is a target in its own right and a target for attackers seeking access to its customers. A typical end-user organization can absorb a breach and recover. For Span, the response itself becomes the product on display. Englman said the company has to be able to demonstrate that controls were in place, that the failure was contained, and that the incident was handled with the same discipline it offers customers. Reputation is what gets sold, and negligence would end the business.
This threat model forces Span to maintain a higher standard of transparency and accountability than many of its clients. Every security incident, even a minor one, must be documented and communicated in a way that reassures customers. The company regularly publishes post-incident reviews and engages in third-party audits to validate its security posture. Englman also highlighted that the nature of IT services means Span often handles sensitive customer data, making data protection and privacy compliance paramount. The regulatory landscape, including GDPR and sector-specific frameworks, adds additional layers of complexity.
Skills Shortage, Restated
The widely discussed cybersecurity talent gap, in Englman’s view, is misframed. Entry-level applicants are abundant. Senior practitioners with five or more years of operational depth are scarce, and that gap cannot be closed quickly through training programs. The Span Cyber Security Center has trained more than 3,000 people, and Englman said the pipeline matters precisely because the industry’s push toward automated tooling threatens to eliminate the junior roles where future experts get built. His measure for a SOC analyst centers on whether they can explain what the alert means and how the conditions that triggered it came about. Without that understanding, an analyst rolling a fifty-fifty guess on relevance is no better than a model doing the same.
The industry’s focus on automation as a substitute for human expertise risks creating a two-tier workforce: a few highly paid senior analysts who manage the tools, and a large number of unskilled operators who simply triage alerts. Englman argues that this model is unsustainable. Junior hires need exposure to real incidents, messy log parsing, and direct interaction with senior mentors to develop the judgment that machine learning cannot replicate. Security programs that invest in on-the-job training and clear career progression pathways will ultimately outperform those that rely solely on automation.
The Wisdom Discarded
Asked which piece of conventional security wisdom he has stopped believing, Englman named the framing of humans as the weakest link in the chain. He called it lazy and a form of blame culture. The responsibility, he said, sits with the CISO to build systems where a user clicking a malicious link does not bring the environment down. Brittle defenses that depend on perfect human behavior are a design failure. This perspective aligns with a broader shift in security engineering: moving from trying to control user behavior to designing resilient systems that assume users will make mistakes.
In practice, this means implementing defense-in-depth architecture, such as network segmentation, endpoint detection with automated containment, and privilege management that limits blast radius. It also involves fostering a security culture where employees are encouraged to report suspicious activity without fear of reprisal. Englman emphasized that trust is not a weakness; it is a strategic asset. When employees trust the security team to handle incidents without punitive measures, they become active participants in the defense rather than obstacles to overcome.
The evolution of the CISO role in the age of AI is still being written. Leaders like Englman are navigating a landscape where the speed of innovation often outstrips the speed of governance. By focusing on secure enablement, realistic expectations for automation, and system resilience over human perfection, they are defining a new playbook for cybersecurity that balances risk with business agility. As agents become more autonomous and AI assistants more capable, the core challenge remains the same: how to protect an organization without stifling the very creativity and productivity that drive its success.
Source:Help Net Security News