
Refrigeration Controller Cybersecurity: What the CISA/FBI Advisory Says, What Has Happened Since, and a Cold-Chain OT Hardening Checklist


Team Shieldworkz
On November 5, five US agencies put their names on a short fact sheet about a very unglamorous piece of equipment: the controller that runs a cold room. CISA and the FBI were joined by the Department of Health and Human Services, the Department of Agriculture and the Department of Transportation. Their message was blunt. Attackers are going after electronic temperature controllers and commercial refrigeration controllers in retail grocery and cold storage facilities, and the first fix is to get those controllers off the public internet.
The advice reads like basic hygiene. The context is what makes it serious. Over the past three months, refrigeration controllers have been used to spoil food at hundreds of French supermarkets, to push temperature thresholds to extreme values at French blood-plasma storage sites, and, in an investigation that is still open, possibly to disrupt freezers at US military commissaries. In the days since the advisory, France has formally blamed Iran’s IRGC, researchers have published six unpatched pre-authentication flaws in a widely deployed Danfoss controller, and attackers started probing newly disclosed Copeland flaws within two hours of publication.
This post covers the advisory itself, the latest developments, and a practical hardening checklist written for the people who actually have to carry it out: CISOs, OT security leads and the facilities and engineering teams who own the equipment.
Key takeaways
The joint fact sheet was published on November 5 and lists eight mitigations: eliminate public internet exposure, isolate OT devices, enforce credential security, apply patches, restrict access, keep reliable backups, monitor and report, and review setpoint ranges.
CISA’s Chris Butera said every organization that depends on refrigeration and monitoring systems should assume it could be a target. The fact sheet aligns with the Cross-Sector Cybersecurity Performance Goals from CISA and NIST.
On November 9, France attributed the October attacks to Shahid Kaveh, an IRGC cyber unit. Other analysts, including SentinelOne, say attribution across the wider wave of incidents remains open.
Exposure is falling, but mostly because people are pulling plugs, not patching. Censys saw global exposure of the two named product families drop 53% between October 9 and November 5, and found patched and unpatched consoles disappearing at the same rate.
Disconnecting is not the finish line. New vulnerabilities in Danfoss and Copeland controllers surfaced within days of the advisory, and some affected hardware is end-of-life.
What the CISA/FBI advisory actually says
The document is a joint fact sheet titled “US Government Partners Urge Hardening Industrial Refrigeration and Cold Chain Monitoring Systems.” Four points matter most for planning.
Scope. It covers retail grocery and cold storage, but it names four critical infrastructure sectors that rely on these systems: Healthcare and Public Health, Chemical, Food and Agriculture, and Transportation Systems, including refrigerated transport. These systems handle automated and remote monitoring, continuous logging, defrost management and energy optimization. The agencies describe the consequences as food spoilage, equipment damage and fire hazards.
Tone. The agencies do not describe a theoretical risk. They say malicious actors are targeting vulnerabilities and weaknesses in unpatched or internet-accessible controllers globally. FBI cyber leadership described the activity as something the Bureau and its partners are seeing around the world.
The eight mitigations. As summarized from the fact sheet:
Eliminate public internet exposure. Do not expose control devices or their management systems to the internet. Where remote access is needed, put it behind a firewall or edge device instead of a port forward.
Isolate OT devices. Separate refrigeration systems from office networks and from other sites.
Enforce credential security. Change default passwords and use phishing-resistant multifactor authentication where it is supported.
Apply patches. Keep firmware current and coordinate updates with the equipment supplier.
Restrict access. Limit who and what can reach controllers, and what each account can change.
Keep reliable backups. Maintain tested copies of controller configurations.
Monitor and report. Watch for suspicious activity and report it to CISA’s 24/7 operations center or the FBI’s Internet Crime Complaint Center.
Review setpoint ranges. Know what safe temperature and defrost limits look like so that out-of-range values stand out.
Reference material. The fact sheet points operators to CISA’s guidance on primary mitigations for reducing cyber threats to operational technology and its internet exposure reduction guidance.
Note for readers who need the exact wording: the PDF is on CISA’s site, and anything you build into policy should be checked against the original rather than a summary, including this one.

Why this advisory landed now: the timeline
The advisory is the end of a short, fast chain of events. Here is the sequence as it is currently reported.
August. The Pentagon later said it is treating a series of unexplained freezer malfunctions at US military commissaries as a suspected cyberattack. Reuters reported at least six commissaries had refrigeration failures, and at least one had equipment enter defrost mode and heat food overnight. The investigation remains open and no cause or perpetrator has been established. Also in August, Claroty publicly disclosed critical vulnerabilities in the Danfoss AK-SM 800A and 23 flaws in the Copeland XWEB Pro. Copeland patched the XWEB Pro issues in firmware.
October 10 to 12. Attackers took over refrigeration controllers at hundreds of French supermarkets and independent stores, switching units into defrost mode overnight. France’s finance minister told lawmakers that roughly 700 of the country’s about 10,000 supermarkets were affected. Reported losses run to tens of millions of euros, and some stores had fires.
October 12 to 16. France’s CERT-FR issued an emergency alert naming Danfoss AK-SM and AK-SC series controllers, then added Copeland and Dixell XWEB systems the next day and updated its recommendations on October 16.
November 2 to 3. A second wave hit stores in several European countries, including Sweden, Estonia, Ireland, Luxembourg, the Netherlands and Belgium. Danfoss said its products were involved, and Copeland confirmed that some of its XWEB controllers it had previously flagged as vulnerable were targeted.
November 2. Attackers reached three sites of the French Blood Establishment (EFS) through a refrigeration equipment supplier and raised temperature and defrost thresholds to extremely high levels in rooms holding plasma. EFS said monitoring and alarm systems caught the change quickly, products were moved or secured, and no blood products were lost. A French weekly reported that setpoints were pushed as high as 110°C from around minus 30°C; EFS has not confirmed the figure.
November 5. The CISA/FBI joint fact sheet is published.
November 9. France attributes the October attacks to Iran’s IRGC.
November 11 and after. Researchers disclose unpatched Danfoss AK-SM800A flaws, and VulnCheck discloses 14 Copeland XWEB EVO flaws.
The blood bank incident deserves a second look. The attackers did not need to hit the blood service at all. They hit a supplier, and the supplier’s control path reached the cold rooms. That is a third-party OT access problem as much as a controller problem.
Latest developments since the advisory
France blames Iran, but attribution is not settled across the board
On November 9, France’s foreign minister confirmed the attribution of the October 11 to 13 attacks to Shahid Kaveh, an IRGC cyber unit better known under the persona CyberAv3ngers. Paris prosecutors opened a judicial investigation into organized cybercrime committed to serve a foreign power, France summoned Iran’s ambassador, and it said it would seek European sanctions. EU officials expressed solidarity and said they were coordinating a response.
Separately, analysts urge caution. SentinelOne’s assessment says attribution for the wider wave remains open and that the incidents may not share one operator, and the pro-Russian group NoName057(16) and another group called Z-Pentest have published claims of their own. French officials themselves have cautioned against linking every refrigeration incident to one group. For defenders the practical point is the same either way: the entry path was exposed, and a state-linked actor, a hacktivist group or an opportunist can all use it.
The entry path may be simpler than the zero-days
ANSSI’s early findings, as reported, pointed to exposed admin pages and weak passwords rather than the older published vulnerabilities. That matches the fact sheet’s emphasis on exposure and credentials. Danfoss has said most of the affected French units were end-of-life AK-SM 800 and AK-SC 355 models, with some AK-SM 800A units on outdated firmware. Danfoss also told Belgian media the weakness lies mainly in outdated software.
If that picture holds, the lesson is uncomfortable. Many of these attacks did not need a clever exploit. They needed a reachable login page.
Six new pre-authentication flaws in the Danfoss AK-SM800A
Midnight Blue researcher Carlo Meijer reported six pre-authentication remote code execution flaws in the AK-SM800A. They are stack-based buffer overflows affecting every firmware version up to and including 5.3.2, and the web process runs as root. Midnight Blue showed a working root exploit on an older firmware version and is withholding details while it coordinates disclosure with Danfoss. Three things stand out for CISOs:
Danfoss made “strict” mode the default from firmware 4.0, which blocks most of the unauthenticated web attack surface, but the researchers say another overflow in the login-checking component still works on the latest firmware with strict mode on.
The controller does not use the chip’s secure boot features, so an attacker with root could plant an implant that survives a factory reset, or wipe flash and brick the device. That turns a spoiled-food incident into a hardware replacement project, with cold rooms idle for days.
At the time of the latest reporting there was no fixed firmware and no CVE identifier. Germany’s BSI, Belgium’s CCB and Ireland’s NCSC advise taking AK-SM800A management interfaces off the internet, with a VPN as the minimum for any remote connection.
Copeland XWEB EVO: exploitation attempts within two hours
VulnCheck disclosed 14 vulnerabilities in the XWEB EVO, a Dixell supervisory system now sold under the Copeland name, and said each was shown to give root on test systems. Exploitation attempts began in under two hours after publication. VulnCheck was careful to say that this opportunistic probing is not evidence of the October campaign France attributed to the IRGC. Copeland ended software support for the XWEB EVO family on March 16 and told customers to move to XWEB Pro. The takeaway: end-of-life controllers are not just unpatched, they are unpatchable, and the window between disclosure and exploitation has shrunk to hours.
Exposure is dropping, but not for the reason you want
Censys tracked controllers named in the CERT-FR alert and found global exposure fell 53% between October 9 and November 5, from 6,695 to 3,124 hosts. Danfoss StoreView dropped 76% (3,004 to 708) and Copeland/Dixell XWEB dropped 35% (3,708 to 2,420). In France, StoreView consoles fell from 1,825 on October 11 to 537 on October 14. Censys also found that patched and unpatched consoles disappeared at the same rate, which suggests operators were disconnecting rather than updating. Of 220 consoles still exposed in France at that point, 45% ran firmware older than version 4.2.
Zoom out and the numbers are larger. Forescout Research counted 12,976 internet-exposed service observations across 10 product groups, tied to 6,017 distinct public IP addresses in 99 countries and territories, covering Copeland/Dixell, Danfoss, CAREL and other vendors. Scan counts differ between sources because tools count hosts, services and ports differently, so treat any single figure as a snapshot.
Why refrigeration is an OT problem, not a facilities problem
A refrigeration controller looks like building equipment, so it usually sits outside the OT security program. It should not. These devices accept setpoints, schedule defrost cycles and command compressors. They are reachable by technicians, integrators and vendor cloud platforms, often through a 4G modem someone installed years ago. Few vendors make this equipment, so a flaw in one product line can reach thousands of sites at once. And the same attack that spoils groceries can destroy vaccine stock, plasma, reagents or chemical feedstock.
Two structural problems explain most of what we have seen:
Fragmented ownership. Facilities understand the physical process. IT sees the network. The vendor owns the firmware, the integrator owns the modem and the store manager owns nothing. Nobody holds the whole attack path.
Lifecycle debt. Controllers live for a decade or more. When the vendor ends support, a vulnerable box stays in production because replacing it costs money and needs a maintenance window.
The cold-chain OT hardening checklist
The checklist below follows the advisory’s eight mitigations and adds the operational detail that CISOs and OT leads usually have to work out themselves. Items marked (advisory) come directly from the fact sheet’s themes. Items marked (added) are Shieldworkz practitioner guidance and reflect what the incident reporting suggests.
First 24 hours: stop the bleeding
Find every refrigeration and temperature controller you own, including those run by contractors and landlords. Check store, warehouse, lab, pharmacy and transport assets. Include integrator-supplied 4G or cellular modems and vendor cloud portals. (added)
Scan your own public IP ranges and your vendors’ remote access paths from outside. Look specifically for management web interfaces, including StoreView, XWEB and similar consoles. (advisory)
Remove any controller from the public internet. Port forwards, open web ports and “temporary” vendor access are the usual culprits. (advisory)
Change every default and shared password. Retire the credentials printed in the manual. (advisory)
Confirm your independent alarm path works. Alarms that route through the controller you are trying to protect will not warn you when it is compromised. (added)
First 7 days: close the access path
Replace direct exposure with secured remote access. A firewall or edge device with named identities and phishing-resistant MFA where supported. A VPN, private APN or tightly scoped IP allow list is a minimum; CERT-FR describes allow lists as a short-term measure only. (advisory, added detail)
Remove shared and stale contractor accounts. Require named identities, expire access when the work ends, and review who still holds vendor credentials. (added)
Restrict by function, not just by device. An engineer who needs to read an alarm should not be able to change a defrost schedule or setpoint. Where the controller cannot enforce that, put the control in the gateway or access layer. (advisory, added detail)
Segment controllers. Separate VLAN at minimum, and keep them away from point-of-sale systems, staff laptops and other sites. (advisory)
Take offline backups of controller configurations and prove you can restore one. (advisory)
First 30 days: fix the lifecycle
Record firmware version, support status and end-of-life date for every controller. Mark anything out of vendor support as a replacement project with an owner and a date. (added)
Patch in coordination with the supplier. Test in a non-critical location first, and confirm what each update actually fixes. Note that for some recently disclosed flaws no fixed firmware exists yet, so patching alone is not a plan. (advisory, added detail)
Add compensating controls for devices you cannot patch. Gateway enforcement, protocol filtering, no direct reachability and tighter monitoring. (added)
Document and enforce safe setpoint ranges. Define normal temperature and defrost limits per asset class, and where hardware allows, enforce them outside the controller’s own software. (advisory, added detail)
Ask every refrigeration vendor and integrator for their vulnerability disclosure, patch and end-of-life commitments in writing. Include a requirement that they tell you when their own remote access tooling is compromised. (added)
First 90 days: build detection and response
Correlate remote sessions with configuration changes. Alert on setpoint, defrost schedule, alarm threshold, user and network changes, especially outside business hours. (advisory, added detail)
Deploy passive OT monitoring so you can see controller traffic, firmware versions and new connections without touching the devices. (added)
Write a cold-chain incident playbook. Who calls whom at 2 a.m., how to manually override, where product is moved, and who decides. Be careful with automatic isolation: cutting a controller off without understanding the physical process can cause its own damage. (added)
Run a tabletop exercise built on the French scenario. Hundreds of sites change state overnight, the vendor is unreachable, and customers are asking questions by 6 a.m. (added)
Know where to report. CISA’s 24/7 operations center or the FBI’s IC3 in the US; your national CSIRT elsewhere. (advisory)
Five questions to put to your team this week
How many refrigeration, freezer and temperature-monitoring controllers do we own or depend on, and who is the named owner of each?
Can any of them be reached from the internet, directly or through a vendor’s modem or cloud console?
Which of them are out of vendor support?
If a controller reported a safe temperature while the room was warming, how would we know?
If a supplier’s remote access was hijacked tonight, what could it change in our cold rooms?
If your honest answer to any of these is “we don’t know,” that is the finding.
Frequently asked questions
What is the CISA/FBI refrigeration controller advisory?
It is a joint fact sheet published on November 5 by CISA, the FBI, HHS, USDA and the Department of Transportation. It warns that malicious cyber actors are targeting electronic temperature controllers and commercial refrigeration controllers in retail grocery and cold storage facilities, and it lists mitigations focused on removing internet exposure, strengthening access and monitoring.
Which refrigeration controllers have been attacked?
France’s CERT-FR named Danfoss AK-SM and AK-SC series controllers and Copeland and Dixell XWEB systems. The US fact sheet does not name specific products.
Is a VPN enough to protect a refrigeration controller?
A VPN is a sound minimum compared with a direct internet exposure, but it is not sufficient on its own. You still need named identities, MFA, least privilege, segmentation, patching or compensating controls for unpatched firmware, and monitoring. For end-of-life equipment, the longer-term answer is replacement.
Does this apply outside the United States?
Yes. The agencies describe the activity as global, the incidents so far have centered on Europe, and Forescout’s scans found exposed controllers in 99 countries and territories.
Can attackers cause physical damage through a refrigeration controller?
Yes. Reported effects include food spoilage, equipment damage and fires. Researchers have also warned that a compromised AK-SM800A could be bricked or implanted in a way that survives a reset, which would force hardware replacement.
How Shieldworkz helps cold-chain and critical infrastructure operators
Shieldworkz is an end-to-end industrial OT cybersecurity company covering SCADA, OT and IoT environments with an integrated mix of platform, solutions and services. For cold-chain, food, pharma and healthcare operators, that typically means:
OT asset discovery and exposure assessment to find controllers, modems and remote paths you did not know about.
Risk and vulnerability assessment that tracks firmware, support status and known flaws against each asset.
Segmentation and secure remote access design so vendor and contractor access is identity-based, scoped and logged.
Continuous OT monitoring and detection tuned to configuration and setpoint changes.
Incident response readiness, including cold-chain playbooks and tabletop exercises.
If you want a quick view of your own exposure, talk to the Shieldworkz team about an OT exposure and cold-chain readiness assessment.
Securing Cold-Chain Infrastructure Starts with OT Visibility
The recent attacks on refrigeration controllers demonstrate how exposed OT systems can disrupt operations, damage equipment, and put critical products at risk. The CISA/FBI advisory is a timely reminder that securing industrial environments starts with knowing what is connected, who can access it, and how those systems are protected.
For cold-chain operators, removing internet exposure is only the first step. Strong access controls, network segmentation, vulnerability management, continuous monitoring, and a clear response plan are essential to reducing risk. Legacy and end-of-life controllers also require particular attention, especially when security patches are no longer available.
The priority is clear: move from reactive protection to proactive OT security.
Shieldworkz helps organizations identify hidden OT risks, assess vulnerabilities, strengthen remote access, and improve visibility across connected industrial environments. With the right security measures in place, operators can better protect refrigeration systems, maintain operational continuity, and build long-term resilience against evolving cyber threats.
Don't wait for a refrigeration failure to reveal a cybersecurity gap. Assess your OT exposure before attackers find it.
Additional resources:
Incident Response Plan for Municipal & Wastewater Utilities here
What Is Removable Media? Risks, Policies, and Industrial OT Security Solutions here
Free Removable Media Policy Template for OT and IT Teams here
Remediation Guides here
Wöchentlich erhalten
Ressourcen & Nachrichten
Erfahren Sie, wie unsere branchenführenden OT-Security-Lösungen kritische Sicherheitsherausforderungen gemäß KRITIS-Anforderungen bewältigen
Dies könnte Ihnen auch gefallen.

2027 OT Security Budget Guide: Where to Spend and How to Justify It

Team Shieldworkz

47-Day TLS Certificates and Industrial PKI: How OT Teams Should Prepare

Team Shieldworkz

Microsoft outage to physical Harm: Investigating the Brühl Hotel chlorine incident

Prayukth K V

Investigative cybersecurity report: The TSMC employee data breach and the security implications of cloud and third-party dependencies

Prayukth K V

How to Secure PLCs: 12 Ways to Harden Controllers Without Breaking Production (Updated for the Latest CISA Advisories)

Team Shieldworkz

Cascading failure in the Cloud Core: Investigative assessment of the November 2026 global Microsoft outage and systemic infrastructure risk

Prayukth K V

