Electrical power grid substation

(Photo by ABCDstock on Shutterstock)

Passing Every Cybersecurity Audit May Still Leave Grids Exposed

In A Nutshell

  • A new computer model suggests insiders with legitimate access can reach a power grid’s sensitive controls in fewer steps than outside hackers.
  • That gap comes from how the model treats the two attack types, not from measured real-world break-ins.
  • Mandatory grid cybersecurity standards focus mostly on credentials and training, not on stress, grievances, or fatigue among workers with access.
  • In simulations, no single defense worked best against every attacker mix, so the authors favor layered protections.

Power utilities can be fined up to $1.5 million per violation per day for breaking mandatory cybersecurity rules. Yet a new paper argues those rules can miss the people who already hold the keys. Its mathematical model suggests that a worker or contractor with legitimate access could reach a grid’s sensitive controls in fewer steps, and with better odds of success, than a hacker working from outside.

Researchers at Michigan Technological University and the University of Michigan-Dearborn built the model to compare how outside hackers and trusted insiders move through the control systems behind America’s electrical grid. Their paper appears in the journal Cyber-Physical Energy Systems.

Those systems let operators monitor and adjust substations, breakers, and transmission lines remotely from a control room. Long presumed to be sealed off, substations and control centers are often not fully isolated, the paper notes, with openings through laptops, removable media, contractor devices, and remote connections.

A Sewage Spill Shows Insider Threats Come Through Trusted Access

In early 2000, Vitek Boden, a former contractor who had helped install a wastewater control system in Queensland, Australia, used stolen radio equipment and inside knowledge to send unauthorized commands to remote pumping stations. Pumps and valves began operating improperly, and roughly 800,000 liters of untreated sewage spilled into local waterways and surrounding land. The attack did not come from a conventional outside breach of technical protections. It came through trusted access.

That incident happened outside the power sector, years before today’s U.S. grid cybersecurity standards existed. The researchers still see the same pattern in the electricity sector: insider misuse, disgruntled employees, vendor compromise, and phishing emails that trick trusted staff into handing over access. Many cases go unreported over reputation and regulatory worries, so the true scale of insider risk is hard to pin down.

Power Grid Cybersecurity Infographic
A new model suggests insiders can reach power grid controls in fewer steps than hackers, and audits may not catch the risk. (Image by StudyFinds)

Insider Threats Need Fewer Steps Than Outside Hackers, Model Suggests

Outside attackers generally have to work for it. Breaking into grid controls usually means a chain of difficult moves: scanning networks for weaknesses, tricking staff into giving up passwords, getting past firewalls, and working from a company’s everyday business network into the separate system that runs equipment. Each stage carries a chance of failure or detection.

Insiders skip most of that. Someone with valid credentials and inside knowledge can head almost straight for sensitive controls, with no need to guess passwords or trick a coworker into clicking a bad link. In the model, that shortcut concentrates risk around the outcomes with the biggest consequences. That gap comes from how the model treats the two kinds of attacks, not from measuring real break-ins.

Motives vary. The paper lists patterns from earlier research: employees facing demotion or pay disputes, tired operators whose focus slips during long shifts, contractors with special access and a financial reason to misuse it, and careful workers whose mistakes give an outside attacker a foothold. In many cases, the paper says, the most severe scenarios come from outsiders leaning on an insider’s access or knowledge to speed up a break-in.

Simulations Show No Single Defense Beats Every Type of Attacker

To compare the threats on equal footing, the team modeled a grid’s control system as constantly shifting among three conditions: running normally, weakened, and fully compromised. The model calculates how much time the system spends in each over the long run, depending on whether hackers, insiders, or a mix of both are pushing it around. Each simulation ran 10,000 times to smooth out chance swings.

Researchers then tested five defenses against four attacker mixes: mostly outside hackers, mostly insiders, mostly attackers who tamper with vendors’ software, and an even blend. A setup called zero trust segmentation, which limits how far any user can roam inside a network, cut the time the modeled system spent compromised by about 69 percent and came out best in three of the four mixes. It also raised the modeled risk to physical equipment by about 27 percent. Locked-down computers reserved for administrator tasks cut that physical risk by about 46 percent and became the top pick once insiders made up roughly 44 percent or more of attackers. No defense substantially reduced the risk of damage spreading deeper, and several made it worse, so the authors favor layering several protections. Those percentages come from the paper’s illustrative simulations, not from a real utility.

Passing a Cybersecurity Audit Does Not Rule Out Insider Threats

Mandatory standards called NERC CIP, first introduced in 2006, require screening and training for workers with special access. According to the paper, though, the standards stay mostly procedural, centered on verifying credentials rather than checking whether an employee under financial stress, holding a grievance, or running on empty might become a liability. A utility can check every box and still carry hidden risk on the human side.

Grid defense has long centered on keeping hackers out. The model points closer to home. One of the biggest risks to the system that keeps the lights on may not be a stranger probing for weaknesses from halfway around the world. It may be someone who already has a badge, a login, and a reason to be upset. Defenses that ignore that person overlook the same kind of threat that turned a sewage system into a spill site.


Disclaimer: This article summarizes a single research paper for general information. Its findings come from a computer model and simulations, not from real-world power grid data, and should not be treated as security, legal, or professional advice.


Paper Notes

Limitations

Ten and colleagues built an abstract framework that is not tied to any specific network layout or real power system. Detailed physical consequences, such as power-flow violations, equipment-level cascading failures, and unstable grid behavior, are treated as a separate layer that would need network-specific data and fall outside the paper’s scope. Insider threats, stealthy intrusions, and supply-chain compromises are often rare and underreported, which limits how precisely their likelihood can be measured from historical data. To address that, the researchers combined operational data with expert judgment and threat-intelligence information, and they built in methods to track how uncertainty in those estimates affects the final risk numbers. The attacker analysis is an approximation, not a full model of how a real attacker and defender would respond to each other over time. The defense results come from aggregate simulation outputs meant to illustrate the framework, not from a real utility’s systems. The study also does not identify which individual employees might pose a risk.

Funding and Disclosures

No funding sources or grant numbers appear in the paper. Its competing interest declaration states that Junho Hong is an associate editor for Cyber-Physical Energy Systems and was not involved in the editorial review or the decision to publish the article, and that all authors declare no competing interests.

Publication Details

Chee-Wooi Ten and Lawrence Dilworth of the Department of Electrical and Computer Engineering at Michigan Technological University; Junho Hong and Swetha Rani Kasimalla of the Department of Electrical and Computer Engineering at the University of Michigan-Dearborn; and Roman Sidortsov of the Department of Social Sciences at Michigan Technological University wrote the paper, titled “The emerging landscape of cyberthreats in SCADA deployment.” It was published in the journal Cyber-Physical Energy Systems (Volume 2, 2026, pages 1 to 24) after being received on June 7, 2026, and accepted on July 29, 2026. The DOI is 10.1016/j.cpes.2026.07.003. The article is open access under a Creative Commons CC BY license.

About StudyFinds Analysis

Called "brilliant," "fantastic," and "spot on" by scientists and researchers, our acclaimed StudyFinds Analysis articles are created using an exclusive AI-based model with complete human oversight by the StudyFinds Editorial Team. For these articles, we use an unparalleled LLM process across multiple systems to analyze entire journal papers, extract data, and create accurate, accessible content. Our writing and editing team proofreads and polishes each and every article before publishing. With recent studies showing that artificial intelligence can interpret scientific research as well as (or even better) than field experts and specialists, StudyFinds was among the earliest to adopt and test this technology before approving its widespread use on our site. We stand by our practice and continuously update our processes to ensure the very highest level of accuracy. Read our AI Policy (link below) for more information.

Our Editorial Process

StudyFinds publishes digestible, agenda-free, transparent research summaries that are intended to inform the reader as well as stir civil, educated debate. We do not agree nor disagree with any of the studies we post, rather, we encourage our readers to debate the veracity of the findings themselves. All articles published on StudyFinds are vetted by our editors prior to publication and include links back to the source or corresponding journal article, if possible.

Our Editorial Team

Steve Fink

Editor-in-Chief

John Anderer

Associate Editor