Image: dragos.brightspotcdn.com · rights & removal
Industrial Ransomware Analysis for Q3 2026
Reporting by Dragos (OT security)Read the original at dragos.com
Executive Summary
Ransomware activity against industrial organizations increased in the third quarter of 2026, rising to 1,266 incidents, which is an 11% increase over the previous quarter's 1,140 incidents. Manufacturing was the most affected sector, accounting for 876 incidents (69%), followed by the ICS ecosystem with 116 incidents, and transportation and logistics with 105 incidents. The U.S. remained the most impacted country with 473 incidents (37% of the global total).
The threat landscape shows continued reliance on established tactics, including exploiting edge devices, using stolen credentials, and evading endpoint detection and response systems. Specific incidents included a July attack on Fairlife that suspended production across U.S. facilities for approximately 11 days, and an August attack on Boston Scientific halting global manufacturing and shipping for two weeks.
Ransomware groups like Qilin and The Gentlemen claimed significant victim volumes, accounting for 362 incidents (29%) combined, while the number of distinct ransomware brands claiming industrial victims increased from 70 to 99. The extortion model is shifting towards data theft-only operations, and geopolitical influences are noted in activity, including public reports linking state-aligned adversaries to ransomware branding.
The exploitation of vulnerabilities in enterprise software, such as flaws in VMware vCenter and JetBrains TeamCity, was linked to subsequent ransomware operations. Specific campaigns involved exploiting vulnerabilities in SonicWall appliances and credential theft via the FortiBleed campaign, which linked to groups like INC Ransom and Lynx.
Facts Only
* 1,266 ransomware incidents affected industrial organizations in Q3 2026 (July through September), an 11% increase from 1,140 in Q2.
* Manufacturing accounted for 876 incidents (69%) of the global total.
* Engineering firms/integrators supporting the operational technology (OT) ecosystem recorded 116 incidents.
* Transportation and logistics recorded 105 incidents.
* The U.S. had 473 incidents in Q3 2026, representing 37% of the global total.
* Qilin and The Gentlemen accounted for 362 incidents (29%) in Q3 2026.
* Ransomware operators continued to exploit edge devices, enterprise applications, stolen credentials, and evade EDR systems.
* Critical flaws in VMware vCenter and JetBrains TeamCity were exploited shortly after disclosure and later tied to ransomware operations.
* INC Ransom exploited SonicWall SMA 1000 vulnerabilities for initial access.
* The FortiBleed campaign harvested credentials from over 86,000 FortiGate devices in June 2026, linking to INC and Lynx operations.
* Cl0p targeted product lifecycle management (PLM) software, exploiting flaws in PTC Windchill and FlexPLM.
* The Gunra ransomware operation was publicly linked to a group associated with the North Korean government.
* Temporary Activity Thread (TAT)26-60 used Claude Code for AI-assisted intrusions against eight organizations in Q3 2026.
Full Take
The data reveals that the threat to industrial organizations is characterized by high volume, sector concentration, and a consistent reliance on exploiting existing enterprise infrastructure gaps rather than deploying novel, highly complex operational system compromises. The growth rate of incidents across all major regions demonstrates a persistent and global adversary focus, with North America maintaining dominance but Europe and Asia experiencing measurable increases that signal expanding operational reach.
The correlation between specific vulnerability exploitation—such as flaws in VMware vCenter and JetBrains TeamCity—and the subsequent deployment of ransomware indicates a critical failure in securing the IT/OT DMZ environment, suggesting that foundational enterprise security weaknesses are the primary vectors for industrial disruption. The persistence of tactics by groups like Qilin and The Gentlemen, combined with the shift toward data theft-only extortion, suggests an evolution where operational impact is achieved through information leverage rather than direct system control.
The evidence regarding the FortiBleed campaign and the use of AI assistants demonstrates a maturation in adversary capability, moving from manual credential harvesting to automated, large-scale pipeline creation using shared infrastructure exploits. The findings on Cl0p underscore that intellectual property exploitation within PLM systems creates significant downstream risk for industrial customers, emphasizing that attacks against design data directly impact supply chain integrity even if direct operational control remains elusive. The pattern suggests that defensive focus must shift from perimeter defense to securing the trust relationships established through enterprise applications and identity management.
Bridge Questions: If the correlation between IT vulnerabilities and OT exploitation is this consistent, what specific governance or segmentation failures in industrial environments are most reliably exploited by these TTPs? How can defenders develop resilience when the primary vector remains credential and application-layer exploitation rather than zero-day kernel exploits? What long-term shifts in regulatory focus are necessary to address the demonstrated persistence of state-aligned groups leveraging RaaS infrastructure across geopolitical lines?
From the original · Dragos (OT security)
In the third quarter (Q3) of 2026 (July through September), analysis of publicly disclosed victim data and ransomware groups’ postings on data leak sites (DLS) identified 1,266 ransomware incidents affecting industrial organizations worldwide, an 11% increase over the 1,140 incidents recorded in Q2.Read the full story at dragos.com
Sentinel unavailable
The automated check of the source article's wording did not complete, so no result is shown.
This looks only at the wording of the original source article, not at this page's AI-written sections. A small local AI model made this estimate. It has not been checked against known human and machine texts, so treat it as provisional. It cannot show who wrote an article.
