Safeguard
Supply Chain Security•

How Cargo Ships Get Hacked and How Shipping Is Secured

Maritime cyber security: how cargo ships, ports and ship systems get hacked, from AIS and GPS spoofing to ransomware, and how shipping is secured.

Safeguard Security Team
67 min read

Cargo ships and ports get hacked mostly through ordinary IT, not through the bridge. Ransomware and destructive malware have stopped shipping lines, ship managers and container terminals again and again, from the 2017 NotPetya outage at Maersk to the 2023 shutdowns at the Port of Nagoya and DP World Australia. On board, the weak points are satellite terminals with default or hard coded passwords, vendor remote access nobody watches, USB drives, and networks that join the office PCs to the bridge and engine room. Two radio systems that every ship depends on, AIS and GNSS, carry no proof of who sent a signal, so they can be spoofed or jammed. And most of the software on a ship comes from equipment makers and their suppliers, so the fix for almost every documented flaw had to travel down a long supply chain.

That is the short answer. This guide is the long one. It explains what software runs on a modern cargo ship and in a container terminal, and where an attacker can reach it. It covers AIS spoofing research, GNSS jamming and spoofing at sea, ECDIS and NMEA weaknesses found by Pen Test Partners, satellite terminal CVEs, ransomware against shipping lines and ports, terminal operating systems, and vendor remote access. It then turns to the rules that now govern ship cyber security (IMO resolution MSC.428(98) and the IACS unified requirements E26 and E27), hardening checklists for shipowners, ports and equipment makers, and the maritime software supply chain.

It is written from the attacker's point of view for defenders. For each attack you will find how it works and why, what the attacker needs, real documented cases, the impact, how to detect it and how to stop it. You will not find exploit code, tool command lines, transmitter settings or step-by-step instructions. Where a tool or a test is named, it is named as context for authorised testing. Incidents are described by their technical event and business impact only; this guide does not attribute any incident to a country or a government.

A note on the law. Testing ships, ports, satellite terminals or radio systems that you do not own, or have no written permission to test, is illegal in most countries. Transmitting on AIS or GNSS frequencies without a licence is also illegal almost everywhere, and it can put lives at risk. Everything in this guide is for protecting your own vessels and terminals, your organisation's fleet with authorisation, or the products you build. The researchers cited here worked on their own equipment, on vessels whose owners invited them, or through coordinated disclosure.

This post is part of a device security series. The guides on how Wi-Fi gets hacked, how routers get hacked and how laptops get hacked cover the crew networks, satcom routers and Windows machines that most ship incidents start on. Our posts on fabless chip designers and the software supply chain and how a laptop is built explain where the chips and SDKs inside marine electronics come from.

Contents

  1. How does a cargo ship work, and what software does it run?
  2. What is the attack surface of a ship and a port?
  3. What has actually happened to ships and ports?
  4. What does an attacker need to hack a ship?
  5. How is AIS spoofed, and why does it work?
  6. How does GNSS jamming and spoofing affect ships?
  7. How are ECDIS, autopilot and bridge systems attacked?
  8. How are satellite links and shipboard networks attacked?
  9. How are engine room, cargo and ballast systems reached?
  10. How does ransomware hit shipping companies?
  11. How are ports and terminal operating systems attacked?
  12. Why is vendor remote access such a large risk?
  13. How should you weigh these risks?
  14. How do you detect attacks on ships and ports?
  15. What do MSC.428(98), IACS UR E26 and UR E27 require?
  16. Why is maritime security a software supply chain problem?
  17. How should shipowners and ship managers secure their fleets?
  18. How should ports and terminal operators secure their systems?
  19. How should maritime equipment makers build secure products?
  20. How Safeguard helps teams that build maritime software
  21. The takeaway
  22. Frequently asked questions
  23. Sources

How does a cargo ship work, and what software does it run?

A modern cargo ship is a floating set of networks. The bridge runs navigation computers, the engine room runs industrial controllers, cargo and ballast systems run their own computers, and an office network handles mail, maintenance and compliance records. A satellite link joins all of it to the shore, and every piece runs firmware or software from an equipment maker.

It helps to walk the ship from top to bottom, because each area has a different owner, a different supplier and a different way of failing.

The bridge. The officer of the watch navigates with an electronic chart display and information system (ECDIS), a computer that shows official electronic charts with the ship's position overlaid. ECDIS takes position from GNSS receivers (the Global Navigation Satellite System, of which GPS is the best known), heading from the gyrocompass, speed from the log, and targets from radar and AIS. AIS, the Automatic Identification System, is a radio transponder that broadcasts the ship's identity, position, course and speed to every ship and shore station in range. The autopilot can steer to a heading, or in track control mode it follows the route planned in the ECDIS. A voyage data recorder (VDR) records bridge data like an aircraft's black box. Many of these devices talk to each other with NMEA 0183, a plain text serial protocol, or over Ethernet.

The engine room. Propulsion control, the power management system that runs the generators, fuel and lubrication systems, and engine monitoring are operational technology (OT): computers that control physical machinery. They are built from programmable logic controllers (PLCs), small industrial computers that run fixed control programs, with operator screens that often run Windows. Pen Test Partners notes that engine diagnostics, energy management, ballast water treatment and exhaust gas cleaning systems may be connected to shore through dedicated gateways for remote monitoring (Pen Test Partners, 2023).

Cargo and ballast. The loading computer works out the stability and hull strength for the cargo the ship is about to load, and nothing goes aboard until it shows the vessel will be stable (Pen Test Partners, 2026). Ballast pumps and valves keep the ship trimmed and stable as cargo moves. Container ships monitor refrigerated containers, and tankers run cargo pumps, gas detection and emergency shutdown systems.

Business IT. Every ship has an office network: ship to shore mail, the planned maintenance system, the safety management system documents, crew records and hours of rest, electronic record books, and the PC that receives chart and publication updates. Pen Test Partners reports that fleet-wide Windows domains are now common, so ship machines share accounts and management with the shore office.

Crew welfare. Crews live aboard for months. Welfare Wi-Fi, shared PCs and personal phones and laptops are a normal part of ship life and a normal part of the network.

Communications. VSAT (very small aperture terminal) satellite systems, older narrowband satellite terminals, low earth orbit (LEO) satellite internet such as Starlink, and cellular in port carry everything to shore. Pen Test Partners lists VSAT, Starlink and cellular as the off-board links on modern cargo vessels, and observes that "VSAT made a usable connection possible. Starlink changed the speed completely."

Five column diagram of the security zones on a ship, navigation and comms, machinery, cargo and ballast, business IT, and crew and guest, each listing hardware, operating system or firmware, components inside and protocols, with a band explaining IACS UR E26 zone boundary rules

Figure 2. Every zone on a ship runs a full software stack, and most of it comes from someone other than the shipowner.

Two facts about this picture matter for everything that follows. First, almost none of this software is written by the shipowner. It comes from ECDIS makers, engine and automation suppliers, satcom vendors, fleet software providers and their own suppliers. Second, ships last a long time and their equipment is rarely replaced, so old operating systems and old firmware stay in service for years. BIMCO's guidelines describe a ship that lost nearly all of its ECDIS computers at sea in a high traffic area because a manufacturer's software update would not run on the outdated operating system underneath it (BIMCO guidelines v5, 2024).

What is the attack surface of a ship and a port?

A ship's attack surface is every path from outside to a computer on board: radio signals such as AIS and GNSS that anyone can transmit, the satellite and cellular links, the shared fleet domain and shore office, vendor remote access and service laptops, USB media, crew devices, and the links between zones that should be separate.

Security people use "attack surface" for the full set of places an attacker can try to get in. For a ship there are three outside paths and several inside ones.

Attack surface map of a cargo ship showing six onboard zones (bridge, engine room OT, cargo and ballast, business IT, crew welfare, comms rack) and three outside paths in (radio signals, shore office and fleet domain, vendors and visitors), with dashed lines for zone links that are often bridged

Figure 1. Where a cargo ship can be reached. Most real compromises start in business IT or the comms rack.

Radio signals. AIS and GNSS are broadcast systems. They were designed for safety and openness, not authentication, so a receiver cannot tell a genuine signal from a well formed fake. Anyone with a transmitter in range can add false AIS targets or drown GNSS signals.

The comms rack and the shore. Satellite terminals, firewalls and routers face the internet. When they are reachable from it and still use factory passwords, they are the front door. Behind them sits the shore office and, increasingly, a fleet-wide Windows domain that lets one set of credentials work on every ship.

Vendors and visitors. Equipment makers maintain their systems remotely, service technicians bring laptops aboard, bunker surveyors and agents plug in USB drives, and chart updates arrive on removable media or through update boxes. Each is a trusted path that bypasses the firewall.

Links between zones. The ship's zones are supposed to be separate. In practice they are often bridged. Pen Test Partners reported in 2020 that OT systems were "often accessible from the IT systems and vice versa," and that it kept seeing dual-homed PCs (computers with a network connection on both sides of a boundary) bridging networks on a large number of vessels, some added by installers and some by crew (Pen Test Partners, 2020).

A port terminal has a similar shape. A terminal operating system coordinates cranes, yard equipment, gates and trucks, exchanges data with shipping lines and customs, and is maintained remotely by its vendor. We cover it in its own section below.

What has actually happened to ships and ports?

The documented record is mostly shore side. Destructive malware and ransomware have stopped shipping lines, ship managers and port terminals for days at a time, while research has shown that AIS, GNSS, ECDIS and satellite terminals can be manipulated. No publicly reported ransomware incident has been confirmed to have run on a vessel's own computers.

Timeline from 2013 to 2026 of documented maritime cyber events: research and disclosure, incidents, and rules, including UT Austin GNSS spoofing in 2013, AIS research in 2014, NotPetya and MSC.428(98) in 2017, COSCO in 2018, CMA CGM in 2020, Transnet in 2021, DNV ShipManager, Nagoya and DP World Australia in 2023, IACS E26 and E27 in 2024, and regional GNSS jamming in 2026

Figure 5. Fourteen years of maritime cyber research, incidents and rules.

Pen Test Partners compiled the public ransomware record against shipping in September 2026, nine incidents from Maersk in 2017 to North Carolina Ports in 2026, and drew the conclusion quoted above: every one hit shore offices, booking systems or terminals, and none was confirmed on a ship (Pen Test Partners, 2026). They also warn that this says more about past conditions than about ships, because what protected vessels (little IT aboard, slow links, no shared domain) is disappearing. The ransomware and port sections below walk through those incidents.

The research record covers the parts of a ship people worry about most: GNSS spoofing of a yacht (2013), AIS spoofing (2014), ECDIS and NMEA manipulation (2018), satellite terminal flaws (2014 onward), and a steady stream of CVEs in marine equipment. A CVE (Common Vulnerabilities and Exposures) identifier is the public ID number given to a specific, disclosed vulnerability.

What does an attacker need to hack a ship?

For most incidents an attacker needs nothing ship specific: a phishing email, stolen or default credentials, an exposed remote access service, or one infected USB drive. Interfering with AIS or GNSS needs a radio transmitter within range. Steering a ship through its systems needs far more: deep access, knowledge of that vessel's equipment, and a crew that does not notice.

It is worth separating three kinds of attacker, because they need very different things.

Financially motivated criminals use the same tools against a shipping line that they use against any company. They need an initial foothold (phishing, a leaked password, an unpatched internet facing system) and then time to move across a Windows network. Pen Test Partners puts it plainly: to these attackers a ship "is another Windows network." They may not even know they are on a ship.

People abusing radio systems need a transmitter and proximity. AIS research in 2014 used a software defined radio transmitter the researchers built themselves (Trend Micro, 2014). GNSS jamming needs only enough power to drown signals that arrive from orbit very weak. GNSS spoofing that keeps a receiver locked to a false position is harder, but the UT Austin team did it in 2013 with a device the size of a briefcase. Range limits all of these: an attacker affects ships within radio reach, not the world fleet.

Someone trying to change how a ship moves needs much more. They need access to the bridge network or the NMEA links, knowledge of which equipment and software versions the ship carries, and a situation where the crew relies on screens instead of radar, eyes and a second position source. That is why the documented cases of physical manipulation are research demonstrations on cooperating vessels, not incidents.

The practical lesson for defenders is that the cheap attacks are the common ones, and they come through IT. Fix those first, then make sure the crew can detect and survive the harder ones.

How is AIS spoofed, and why does it work?

AIS spoofing is the broadcast of false AIS messages, such as a ship that does not exist, a real ship in the wrong place, or a fake distress beacon. It works because AIS has no message authentication: receivers accept any well formed message on the AIS channels, and the identity number inside it is a claim, not a proof.

AIS became mandatory on most ships over 300 gross tonnage on international voyages under a 2002 IMO SOLAS mandate (Wikipedia, AIS), and it is used for collision avoidance, traffic monitoring, accident investigation, aids to navigation and search and rescue (ACSAC 2014). Each station is identified by its MMSI (Maritime Mobile Service Identity) number. Messages go out on two VHF channels and are picked up by nearby ships, shore stations and satellites. Online tracking sites aggregate them for the public.

Three panel diagram of AIS: how AIS normally flows from GNSS and voyage data through the transponder to ships, shore, satellites and tracking sites; where the trust gap is, since any well formed message is accepted and the MMSI is not proven; and what defenders do, such as cross checking against radar and flagging impossible tracks

Figure 3. AIS trusts whoever is talking.

What the research showed

In 2014 Marco Balduzzi, Alessandro Pasta and Kyle Wilhoit published a security evaluation of AIS, presented at ACSAC (the Annual Computer Security Applications Conference) and released as a Trend Micro white paper. Using a software based AIS transmitter they designed, they showed that their concerns affected "all transponders deployed globally on vessels and other maritime stations like lighthouses, buoys, AIS gateways, vessel traffic services and aircraft involved in SAR operations" (ACSAC 2014).

They grouped the threats into three families (Trend Micro white paper):

  • Spoofing. Creating a valid looking but nonexistent ship, faking a collision course to trigger a closest point of approach (CPA) alarm on a target ship, faking an AIS-SART distress beacon, and broadcasting false weather data.
  • Hijacking. Altering the information a real station sends.
  • Availability disruption. Impersonating a maritime authority to reserve all transmission slots (slot starvation), to order transponders onto other frequencies (frequency hopping), or to delay their transmissions (timing attacks), so that ships stop seeing each other.

They also looked at three popular online AIS providers and found that none vetted sources or authenticated data. In the paper's words, the providers "have no means to authenticate the AIVDM sentence senders." The researchers proposed adding certificate based authentication to the AIS protocol, with maritime authorities acting as certification authorities, which tells you what the protocol lacked.

Why it matters

Impact. A fake target can push a ship to alter course, especially in poor visibility. False positions on tracking sites mislead charterers, insurers and compliance teams. Availability attacks can hide every ship in an area from every other ship's AIS display. None of this changes what radar sees, which is why radar and lookout remain the backstop.

What the attacker needs. A transmitter on the AIS channels and radio range to the target. No access to the victim ship at all.

Detection. Cross check AIS targets against radar returns: a target with no radar echo, or a radar echo with no AIS, deserves attention. Watch for duplicate MMSIs, ships that appear from nowhere, impossible speeds, and tracks over land. Shore side, data quality checks on AIS feeds catch many anomalies.

Protection. Train watchkeepers that AIS is an aid, not a sensor of record. Keep radar overlays on, and log AIS with radar so an anomaly can be reviewed. Companies that rely on AIS data for compliance or chartering decisions should use providers that apply source validation and should not treat a single AIS track as proof of anything. The longer term fix is authenticated messaging, which is what the 2014 researchers proposed.

How does GNSS jamming and spoofing affect ships?

GNSS jamming drowns satellite navigation signals so receivers lose their position; it is loud and usually triggers alarms. GNSS spoofing replaces the real signals with counterfeit ones so receivers report a wrong but plausible position, often with no alarm. Both are now seen at sea at scale, and both are countered by cross checking position against radar, gyro and log.

GNSS signals arrive from orbit and are extremely weak by the time they reach a ship's antenna. That makes them easy to drown. Civil GNSS signals also carry no authentication, so a receiver cannot prove a signal came from a satellite.

Two column comparison of GNSS jamming and spoofing at sea: jamming drowns signals, the receiver loses its fix and alarms sound, seen at regional scale in 2026; spoofing gives a wrong fix often without alarm, as UT Austin showed on a yacht in 2013; with a bottom band of detection signals

Figure 4. Jamming takes the position away. Spoofing replaces it with a wrong one.

Spoofing: the UT Austin yacht demonstration

In June 2013 a team led by Todd Humphreys at the University of Texas at Austin was invited aboard the 213 foot yacht White Rose of Drachs as it sailed from Monaco to Rhodes. About 30 miles off Italy, with the owner's permission, they broadcast counterfeit civil GPS signals from a briefcase sized device toward the yacht's two GPS antennas. The false signals slowly overpowered the real ones until they controlled the navigation system. Each time the crew corrected for an apparent drift, the yacht actually turned slightly. "The ship actually turned and we could all feel it, but the chart display and the crew saw only a straight line," Humphreys said. The yacht ended up on a parallel track hundreds of metres from its intended one, and the university noted that unlike jamming, "spoofing triggers no alarms on the ship's navigation equipment" (UT Austin News, 2013).

Jamming at regional scale

Jamming is now routine in several busy waters. On 1 March 2026 maritime analytics firm Windward reported that more than 1,100 vessels experienced GPS and AIS interference in the Middle East Gulf within 24 hours. Ships' AIS positions were shown at airports, at a nuclear power plant and on land, with "crop-circle-type patterns" in surrounding waters, and Windward identified at least 21 new interference clusters. It warned that false locations "create major problems in congested waters where precise navigation is essential for collision avoidance," and that diverted positions created false compliance alerts for banks, charterers and insurers (Windward, 2026).

Notice the link between the two systems. AIS reports whatever position the ship's GNSS receiver gives it, so GNSS interference shows up instantly as bad AIS data for every affected ship.

Impact, detection and protection

Impact. Jamming degrades navigation and AIS across an area and forces ships onto backup methods. Spoofing is more dangerous for one ship, because autopilot in track control will faithfully follow a false position, and an ECDIS will show the ship on its planned line when it is not.

Detection. Compare the GNSS position with radar ranges and bearings to fixed marks, speed over ground with log speed, and course over ground with gyro heading. Watch for many ships at once showing jumps or circles. Modern receivers can report unusual signal strength, or many satellites arriving at the same power, which is a spoofing tell.

Protection. Use receivers that support multiple constellations and frequencies and report integrity alarms. Feed ECDIS from more than one independent position source where possible. Keep radar and visual navigation skills current and drill navigating without GNSS, including manual steering off track control in confined waters. Record bridge data so an event can be reconstructed.

How are ECDIS, autopilot and bridge systems attacked?

Bridge systems are attacked through the computers and serial links they run on. Researchers have found ECDIS units on old, unsupported operating systems with weakly protected configuration pages, and NMEA 0183 messages that carry no authentication at all. Real incidents so far are malware infections and failed updates that took ECDIS offline, not attackers steering ships.

Pen Test Partners' 2018 research

In June 2018 Pen Test Partners presented maritime research at Infosecurity Europe. They tested over 20 ECDIS units and found that most ran old operating systems. On one unit, a poorly protected configuration interface let them change the assumed mounting position of the GPS receiver, which made the ship appear to jump across Dover Harbour. That is not GPS spoofing: the receiver was fine, but the computer's idea of where the antenna sat on the hull was wrong. They could also make the ship appear to be a square one kilometre on each side (Pen Test Partners, 2018).

Because ECDIS often feeds the AIS transponder, a falsified size or position would be broadcast to every nearby ship, which could trigger collision alerts and push other traffic around. And because many vessels run the autopilot in track control, following the ECDIS route, the researchers wrote: "Hack the ECDIS and you may be able to crash the ship, particularly in fog."

The same research looked at NMEA 0183, the serial protocol that carries position, heading and autopilot commands between bridge devices. "There is no message authentication, encryption or validation of these messages," they wrote. A device in the middle of a serial link could alter messages gradually, nudging a ship rather than jumping it. NMEA 0183 has a simple checksum, but a checksum only catches accidental corruption; anyone altering a message can recalculate it.

What the attacker needs. Access to the bridge network or to the serial wiring, which in practice means either physical access or a path from a compromised machine on a bridged network. This is the hard part, and it is why these remain demonstrations.

What real bridge incidents look like

The BIMCO guidelines, written by shipping industry associations, include several anonymised bridge incidents that show the more common failure modes (BIMCO guidelines v5):

  • A newbuild dry bulk ship designed for paperless navigation, with no paper charts aboard, was delayed for several days because both ECDIS networks were infected with a virus. The crew saw it as a technical fault, not a cyber issue. The delay and repairs cost hundreds of thousands of dollars.
  • A ship under pilotage lost its ECDIS and voyage performance computers. The pilot and master navigated visually and by radar. The computers ran outdated, unsupported operating systems and had been failing repeatedly.
  • A ship lost nearly all navigation systems at sea in a high traffic area and reduced visibility after a manufacturer's software update would not run on the outdated operating system, and navigated by one radar and paper charts for two days.

None of these needed an attacker. They show what an attacker would exploit: old operating systems, updates delivered without testing, and crews that cannot tell a cyber problem from a hardware one.

Chart update paths and connected bridge devices

ECDIS needs regular chart and publication updates, so there is always some path from shore to the bridge: a USB stick, a dedicated update PC on the business network, or a connected box. One such box got a CVE in 2026. CVE-2026-2754 describes missing authentication on HTTP API endpoints of NAVTOR NavBox, which let an unauthenticated attacker with network access retrieve internal network parameters "including ECDIS and OT Information, device identifiers, and service status logs." Cydome, which reported it, lists version 4.16.2.4 (November 2025) and later as patched (Cydome advisory). Information disclosure like this is a reconnaissance gift: it tells an attacker what is on the bridge network and how it is addressed.

Voyage data recorders have had CVEs too. CVE-2016-9339 is a path traversal flaw in the INTERSCHALT VDR G4e (versions 5.220 and earlier) that could let an attacker read files on the system.

Detection and protection

Detection. Treat any unexplained ECDIS crash, slowdown or position jump as a possible cyber event and report it. Compare ECDIS position with radar overlay continuously. Keep logs from bridge computers where the equipment supports it.

Protection. Keep ECDIS and the operating system under it within the manufacturer's support period, and test updates before installing them. Lock down configuration pages with strong, unique passwords. Disable unused USB ports and route chart updates through a controlled, scanned path. Keep bridge networks in their own zone, separate from machinery and cargo, as IACS UR E26 requires. Carry the paper or backup arrangements your flag state accepts, and drill using them.

How are satellite links and shipboard networks attacked?

Satellite terminals are the ship's internet edge, and many have been found reachable from the internet with default or hard coded passwords, unsigned firmware and old web interfaces. Once in, an attacker can often reach the rest of the ship, because networks behind the terminal are poorly separated. Crew networks and onboard mail servers add more entry points.

Exposed terminals and default passwords

Pen Test Partners reported in 2018 that "many satcom terminals on ships are available on the public internet" and that many "have default credentials, admin/1234 being very common." They linked terminal version data to GPS positions to map vulnerable ships, then deliberately left that data stale. On a Cobham (Thrane & Thrane) Fleet One terminal they found admin interfaces over Telnet and HTTP, firmware checked only by a CRC (an error check, not a signature), no firmware rollback protection, and admin passwords stored as unsalted MD5 hashes (Pen Test Partners, 2018).

Satellite terminals have a long CVE history. Some examples, with the wording from the CVE records:

CVEProductWhat it is
CVE-2013-7180Cobham SAILOR 900 VSAT, FleetBroadband 150, 250, 500 and othersPassword recovery not properly restricted, allowing administrative access with physical or terminal access
CVE-2014-2940Cobham Sailor 900 and 6000Hard coded credentials for the administrator account
CVE-2018-19393Cobham Sailor 800 and 900Arbitrary writing to the system configuration file
CVE-2016-6551Intellian t-Series and v-Series antennasNon-random default credentials, giving a remote attacker elevated access
CVE-2020-8000Intellian Aptus Web 1.24Hard coded password
CVE-2019-17269Intellian Remote Access 3.18Remote OS command execution through a diagnostic field
CVE-2018-16590FURUNO FELCOM 250 and 500Authentication done only in client side JavaScript
CVE-2018-16705FURUNO FELCOM 250 and 500Unauthenticated access to a file holding all usernames and password hashes
CVE-2026-38056ST Engineering iDirect iQ200 VSAT terminalLocal privilege escalation from a factory configured low privilege technician account

The pattern is consistent across a decade: credentials built into firmware, authentication that can be bypassed, and diagnostic features that run commands. Ruben Santamarta's 2018 research for IOActive found the same thing in maritime antenna controllers: hard coded and in some cases undocumented credentials in Intellian firmware, which he obtained because the vendor's firmware download buckets were openly readable (IOActive, 2018).

Shipboard mail and crew services

In October 2017 IOActive disclosed two critical vulnerabilities in AmosConnect 8, a shipboard email and crew internet platform: a blind SQL injection in the login form that exposed credentials stored in plain text, and a built in backdoor account with full system privileges. There was no fix, because the vendor had announced end of life for AmosConnect 8 in June 2017 (IOActive, 2017). That last detail is the important one for shipowners: software that reaches end of life on shore keeps running on ships for years.

Crew Wi-Fi and the faster link

Crew networks are where policy meets human behaviour. Pen Test Partners describes crew trying to reach bridge networks for unlimited internet after using up their allowance, and a Wi-Fi access point added by a contractor, protected by the contractor's company name, that bridged critical safety systems (Pen Test Partners, 2020). Our guides on Wi-Fi and routers cover the controls in depth.

LEO satellite internet has made ships fast, always connected branch offices. That is good for crews and operations, and it removes the old protection that slow, expensive links gave against large data theft and fast ransomware spread.

Detection and protection

Detection. Scan your own fleet's public IP ranges for exposed management interfaces. Alert on new devices and new outbound connections at the ship firewall. Watch satcom terminal logs for logins from unexpected sources.

Protection. Change every default password at installation and record it in a vault. Never expose terminal management interfaces to the internet; manage them over a VPN or the provider's management network. Keep terminal firmware current and retire unsupported products. Put crew and guest Wi-Fi in its own zone with no route to OT, and make sure no IP address in an OT zone is reachable from outside, which IACS UR E26 now requires for new ships.

How are engine room, cargo and ballast systems reached?

Engine room, cargo and ballast systems are usually reached sideways, not directly. Attackers or malware get onto IT, then cross into OT through dual-homed PCs, shared data historians, vendor remote access or USB media. The documented incidents are infections carried in by people, plus OT systems that crews and installers connected to IT for convenience.

OT systems were historically isolated, so many were never built to defend themselves. Their screens often run Windows, their controllers trust any command on their network, and their software is updated rarely because a change needs the maker, the class society and a port stay.

How the bridges between IT and OT form

Pen Test Partners' field reports describe the mechanisms. On an offshore rig, a data historian (a server that collects readings from industrial systems) had its operator interface on the IT network, which created a bridge into OT. Dual-homed PCs bridging networks were common, "some of these were done intentionally by installers, others clearly accidentally, others deliberately bridged by crew afterwards" (Pen Test Partners, 2020). They also noted a "distinct lack of understanding and interaction between IT and OT installers/engineers on board and in the yard."

What real OT incidents look like

Two BIMCO cases show how infections arrive (BIMCO guidelines v5):

  • A recently built ship had a power management system designed to connect to the internet for updates and remote diagnostics, but it had not been connected. Before connecting it, the company's IT team scanned it and found a dormant worm that could have activated once online. It spread through USB devices, and the owner believed a service technician had brought it aboard.
  • A bunker surveyor asked to print documents from a computer in the engine control room, plugged in a USB drive, and introduced malware onto the administrative network. It went unnoticed until a later cyber assessment.

The loading computer and why "non-essential" IT matters

Pen Test Partners makes a point that ship operators should take seriously. Class rules and IACS UR E22 (which sorts computer based systems by the consequence of failure) put ship IT at the bottom, so it gets little or no design review. Yet a ship cannot load cargo without its loading computer, cannot satisfy a port state control inspection without its safety management records, and cannot run its day without ship to shore mail. Loss of those systems does not sink a ship. It stops it trading (Pen Test Partners, 2026).

Detection and protection

Detection. Passive network monitoring in OT zones, which watches traffic without sending any, can spot new devices, new connections and unusual commands. Periodic scans of OT computers by people who understand the equipment, as in the BIMCO worm case, catch dormant infections.

Protection. Group OT systems into security zones and firewall them, as IACS UR E26 requires: safety systems physically segmented, navigation and comms kept apart from machinery and cargo, and only documented traffic crossing a boundary. Remove dual-homed PCs. Control USB ports and removable media, including those of visitors, and scan media on a dedicated station before it touches OT. Test OT carefully: Pen Test Partners' guidance on testing cargo vessels is built around not putting OT at risk (Pen Test Partners, 2023).

How does ransomware hit shipping companies?

Ransomware hits shipping companies through ordinary corporate IT: phishing, stolen credentials, exposed remote services and compromised software updates. It then spreads across shared Windows domains to booking, EDI, email and terminal systems. Shipping lines have kept ships moving through these events, but at large cost, and fleet-wide domains now give ransomware a path to every vessel at once.

The Maersk business case: NotPetya, June 2017

On 27 June 2017 a destructive malware outbreak, later called NotPetya, spread from a compromised software update. Cisco Talos found that attackers had used stolen administrator credentials to take over the update server of M.E.Doc, a widely used accounting package, and that "all installations came through the M.E.Doc update system." The payload was disguised as ransomware but was destructive, with no way to decrypt the files (Cisco Talos, 2017). Organisations that ran the software, including multinationals with local operations, were hit and the malware then spread across their internal networks.

A.P. Moller Maersk was one of them. The company said the attack "mainly impacted Maersk Line, APM Terminals and Damco," that business volumes were affected for a couple of weeks in July, and that it expected the attack to "impact results negatively by USD 200-300m." Maersk Line could take bookings from existing customers two days after the attack, and the company said it did not lose third party data (CNBC, 2017).

At the World Economic Forum in January 2018, Maersk chairman Jim Hagemann Snabe described the recovery: "We had to install 4,000 new servers, 45,000 new PCs, 2,500 applications." He added that the company saw only a 20% drop in volume and handled 80% of its volume manually during the rebuild, over about ten days (BleepingComputer, 2018).

Three lessons from Maersk apply to every shipping company:

  1. The entry point was a third party software update. Maersk did nothing unusual. A trusted vendor's update channel was compromised. This is a supply chain problem, which is why it sits at the centre of this guide.
  2. A flat, shared network turned one infection into a global outage. Once inside, the malware spread across the estate.
  3. Recovery depended on rebuilding from scratch and on people working by hand. Backups, rebuild capacity and manual procedures decided how fast the business came back.

Later shipping line and ship manager incidents

  • COSCO, July 2018. A ransomware infection hit COSCO Shipping Lines' network in the Americas. The company said local email and network telephones could not work and that it shut down connections with other regions as a precaution, and it warned staff elsewhere not to open suspicious emails (BleepingComputer, 2018).
  • CMA CGM, September 2020. The group reported a cyberattack affecting peripheral servers on 28 September, restored its e-commerce sites (bookings, tracking, invoices) by mid October, and said it suspected a data breach (CMA CGM, 2020).
  • MSC, April 2020. Website and booking tools were affected after an outage at a Geneva data centre (Pen Test Partners, 2026).
  • DNV ShipManager, January 2023. Ransomware hit the servers behind DNV's ShipManager fleet management software on 7 January 2023. DNV said 70 customers operating around 1,000 vessels were affected, that ships could still use the onboard, offline functions, and that no other onboard systems were impacted (TechCrunch, 2023; DNV, 2023). It is a clear example of a software supplier's incident landing on its customers' fleets.

BIMCO adds two anonymised cases: ransomware brought into a shipowner's business network through phishing emails from two unwitting ship agents, after which the owner paid in one case; and ransomware on a ship's main application server, caused by a weak password policy that let attackers guess their way into remote management services (BIMCO guidelines v5).

Why the fleet is now in range

Three stage diagram of ransomware spreading from one compromised machine on one ship, through the fleet-wide Windows domain, to every vessel and the shore office, with the shipboard systems that stop and the controls that break the chain

Figure 6. From one laptop to every ship: why a fleet-wide domain changes the size of a ransomware incident.

Pen Test Partners' September 2026 analysis explains why the absence of vessel ransomware so far is not reassuring. Ships now carry far more IT, that IT replicates to shore, fleet-wide Windows domains are common, and links are fast. "A vessel now looks far closer to a branch office than to an isolated site." They report that across dozens of tests of merchant fleets, "in all but one instance, starting from a single machine on one ship with an ordinary user account, we finished with every Windows machine and every server across the entire fleet." Fleets with segmentation between vessels, firewalling and monitoring slowed them down, and in one case stopped them (Pen Test Partners, 2026).

The impact of that scenario is commercial before it is physical. Without the loading computer, cargo planning falls back to the stability booklet and hand calculation. Without the safety management and maintenance systems, port state control inspections and tanker vetting become very difficult. Without ship to shore mail, every exchange with agents and charterers becomes a phone call. Across a whole fleet at once, with the office down too, that is a business stoppage.

Detection and protection

Detection. Endpoint detection on shore and ship Windows machines, alerts on new domain administrator accounts and on mass changes pushed through group policy or management tools, and monitoring of backup servers, which attackers target first.

Protection. Pen Test Partners' recommendations are specific: tiered administration, no excessive domain administrators, unique local administrator passwords, multifactor authentication on the accounts that matter, local backups on each vessel, and shoreside backups isolated from the Windows domain. Add segmentation between vessels so one ship's compromise does not reach the next, and rehearse running the fleet with office IT down.

How are ports and terminal operating systems attacked?

Ports are attacked the same way as shipping companies, through IT, but the effect is sharper because a terminal operating system coordinates every crane, gate and yard move. When it or the network under it stops, the terminal stops. Nagoya, DP World Australia and Transnet all halted container operations for days.

A terminal operating system (TOS) is the software that runs a container terminal. It holds the vessel stowage plans, the position of every container in the yard, and the queue of moves for every crane and truck. Kaleris, which develops the N4 TOS, says it works with hundreds of terminals and ports worldwide (Kaleris). Around the TOS sit gate automation with optical character recognition (OCR) cameras and truck appointment systems, EDI (electronic data interchange) links that exchange stowage and booking messages with shipping lines, port community systems that share data with customs and hauliers, and the OT that runs cranes and refrigerated container power.

Diagram of a container terminal with the terminal operating system at the centre, connected to vessel and berth planning, EDI with shipping lines, port community and customs systems, quay cranes and yard equipment, gate automation, yard management, the business network and remote vendor support

Figure 7. One terminal operating system coordinates everything, so it is also a single point of failure.

Documented port incidents

  • Transnet, July 2021. South Africa's Transnet halted operations at container terminals in Durban, Ngqura, Port Elizabeth and Cape Town after what it told customers was "an act of cyberattack, security intrusion and sabotage," and declared force majeure. Durban handles over 60% of South Africa's container traffic (CNBC, 2021).
  • Port of Lisbon, December 2022. Website and internal computer systems were affected (Pen Test Partners, 2026).
  • Port of Nagoya, July 2023. A ransomware attack early on 4 July 2023 caused a failure of the Nagoya Port Unified Terminal System (NUTS), the central system controlling all of the port's container terminals. Container loading and unloading using trailers was cancelled until the system was restored. The port handles roughly 10% of Japan's total trade volume (BleepingComputer, 2023).
  • DP World Australia, November 2023. DP World disconnected its Australian ports from the internet to stop "any ongoing unauthorised access," which halted operations at its terminals in Melbourne, Sydney, Brisbane and Perth from Friday to Monday. Trucks could not move containers in or out, and the company handles around 40% of goods entering and leaving the country (BBC, 2023). Pen Test Partners puts the backlog at about 30,000 containers, and DP World later confirmed that employee data had been stolen (ABC News, 2023).
  • North Carolina Ports, August 2026. Gates at three facilities were worked by hand (Pen Test Partners, 2026).

Why ports are hit hard

Impact. A terminal is a just in time system. A two day stop backs up ships at anchor, trucks at the gate and cargo across inland supply chains, and refrigerated and hazardous cargo need special handling while systems are down. DP World's decision to disconnect from the internet was a deliberate containment step: it stopped the attacker but also stopped the business.

What the attacker needs. The same footholds as anywhere else, plus reach into the network that the TOS and its database depend on. Remote vendor support paths and third party integrations (EDI partners, port community systems) widen that reach.

Detection. Monitor identity systems, remote access gateways and the TOS database servers. Watch EDI and portal accounts for unusual activity, because attackers who gain partner accounts can see or alter container data.

Protection. Treat the TOS, its database and the identity system it relies on as crown jewels: segmented, backed up offline, and restorable within a known time. Segment crane, gate and reefer OT from corporate IT. Keep a manual gate and yard fallback and rehearse it, because every port incident above ended with people working by hand.

Why is vendor remote access such a large risk?

Vendor remote access is risky because it is a trusted, privileged path into ship and terminal systems that bypasses most controls. Installers leave remote tools running, vendors share accounts and passwords, and sessions happen without the crew knowing. IACS UR E26 now sets explicit rules: onboard approval, multifactor login, encrypted tunnels, session limits and logging.

Equipment makers need to maintain what they sell. A main engine maker wants telemetry, an ECDIS maker pushes updates, a TOS vendor supports the terminal's database, and a satcom provider manages the terminal. BIMCO lists the systems that commonly have remote access: bridge and engine room computers, reefer monitoring, stability decision support, hull stress monitoring, ECDIS and VDR, load planning, engine monitoring and control, and CCTV. It says the "extent and nature of connectivity of equipment should be known by the shipowner or operator" and that all remote access occurrences "should be recorded for review" (BIMCO guidelines v5).

In practice, Pen Test Partners lists remote management tools among the routes attackers use to move across fleets, and describes maritime technology vendors' approach to security as "variable." BIMCO's ransomware case above began with brute forced remote management services.

What UR E26 and E27 now require

IACS UR E26 section 4.2.6 sets rules for remote access to systems in its scope. Systems must be able to terminate a connection from the ship's end, and "any remote access shall not be possible until explicitly accepted by a responsible role on board." Systems must log all remote access events and keep the logs long enough for review after an incident. "No IP address shall be exposed to untrusted networks," and communication with untrusted networks needs secure connections such as tunnels with endpoint authentication and encryption. For remote maintenance, remote access requires multifactor authentication, a limit on failed logins and a notification message before the session (IACS UR E26 Rev.1).

UR E27 puts matching capabilities into the equipment. Its items 31 to 41 apply to systems that talk to untrusted networks, including multifactor authentication for human users, limits on failed logins, a system use notification, denial of access "unless explicitly approved by authorized personnel on board," remote session termination and cryptographic integrity protection (IACS UR E27 Rev.1).

Six step flow of vendor remote access done properly, request, secure tunnel, onboard approval, MFA and notice, scoped session and logging, with the bad patterns found on real ships and questions to ask every vendor

Figure 8. Vendor remote access, done the way UR E26 and E27 describe.

Detection and protection

Detection. Inventory every remote access path per ship and per terminal, including tools vendors installed years ago. Alert on remote sessions that were not approved through your process.

Protection. Route all vendor access through one controlled gateway with multifactor authentication and per vendor accounts. Require onboard or on site approval for each session, limit it to the system being serviced, and record it. Remove always-on remote desktop tools. Put remote access terms in vendor contracts, and ask suppliers whether their products meet UR E27, including the extra items for untrusted networks.

How should you weigh these risks?

Rank maritime cyber risks by impact and by how cheap the attack is. Ransomware through shore and fleet IT, exposed satcom terminals, unmanaged vendor access and infected USB media are cheap and damaging, so fix them first. Targeted GNSS spoofing, NMEA tampering and supplier compromise are harder but serious, so plan and monitor for them.

Two by two matrix of maritime attack classes by impact and attacker effort: fix first (ransomware on shore, terminals and fleet domains, exposed satcom, vendor remote access, infected USB), plan and monitor (supplier update compromise, targeted GNSS spoofing, ECDIS or NMEA tampering, fleet SaaS compromise), cheap but narrower (AIS fake targets, regional GNSS jamming, crew Wi-Fi bridging) and specialist (AIS availability attacks, VDR tampering, physical bridge access)

Figure 9. Risk versus attacker effort, Safeguard's qualitative reading of the public record.

This is a judgement, not a scoring standard, and your own vessels, trades and equipment will move items around. A tanker operator trading in waters with heavy GNSS interference should rank jamming higher. A container terminal should treat its TOS and identity systems as its top risk. The point of the matrix is to stop scarce budget going to the dramatic scenario while the satcom terminal still has its factory password.

How do you detect attacks on ships and ports?

Detection on ships combines watchkeeping and IT monitoring. On the bridge, compare GNSS and AIS against radar, gyro and log. In IT and OT, watch for new devices, new connections, unapproved remote sessions, new admin accounts and backup tampering. Crews must also be able to recognise a computer fault as a possible cyber event.

Signals for bridge teams

  • GNSS position disagreeing with radar fixes or visual bearings, and speed over ground disagreeing with log speed.
  • AIS targets without radar echoes, duplicate MMSIs, ships appearing from nowhere or on impossible tracks.
  • ECDIS crashes, slowdowns or position jumps, which BIMCO's cases show can be malware.

Signals for ship and fleet IT

  • New devices on any ship network, especially Wi-Fi access points and dual-homed PCs.
  • Remote sessions that were not approved on board, and logins to satcom terminals from unexpected sources.
  • New domain administrator accounts, mass software deployment through management tools, and access to backup servers.
  • USB devices on bridge and engine control room machines outside an approved process.

Signals for ports and terminals

  • Unusual activity on TOS database servers, identity systems and remote access gateways.
  • Unexpected EDI traffic or portal logins from partners.
  • Changes on crane, gate and reefer OT networks.

Make the crew part of detection

BIMCO's ECDIS virus case is the warning: the crew saw a technical fault, not a cyber event, and the ship sat in port for days. Train crews to report unexplained failures, give them a simple route to do so, and make sure the shore team investigates. IACS UR E26 requires an incident response plan for new ships, including procedures to isolate security zones without losing primary functions.

What do MSC.428(98), IACS UR E26 and UR E27 require?

MSC.428(98) requires shipowners to address cyber risk in their safety management system, checked at audits since 2021. IACS UR E26 sets cyber resilience requirements for the whole ship, and UR E27 for the systems and equipment on it, for ships contracted for construction on or after 1 July 2024.

IMO resolution MSC.428(98)

The IMO's Maritime Safety Committee adopted resolution MSC.428(98), "Maritime Cyber Risk Management in Safety Management Systems," on 16 June 2017. It affirms that "an approved safety management system should take into account cyber risk management in accordance with the objectives and functional requirements of the ISM Code," and encourages flag administrations to ensure cyber risks are addressed "no later than the first annual verification of the company's Document of Compliance after 1 January 2021." It points to the guidelines in MSC-FAL.1/Circ.3 for how (IMO, MSC.428(98)). The ISM Code is the International Safety Management Code, which requires every shipping company to run an audited safety management system (SMS). DNV, as a classification society, noted that cyber security would be a mandatory focus area in 2021 annual Document of Compliance audits (DNV, 2020).

In plain terms, MSC.428(98) applies to every company that runs ships under the ISM Code, and it is a management system requirement, not a technical standard. It asks you to identify, assess and manage cyber risk the same way you manage any other safety risk.

IACS UR E26 and UR E27

IACS, the International Association of Classification Societies, published the original E26 and E27 in April 2022, then revised them. Rev.1 of E27 was published in September 2023 and Rev.1 of E26 in November 2023, and only the Rev.1 versions entered into force, on 1 July 2024. They apply to ships contracted for construction on or after that date, including passenger ships and cargo ships of 500 GT and upwards on international voyages, and may be used as non-mandatory guidance for other ships (ABS, 2023).

UR E26: Cyber resilience of shipsUR E27: Cyber resilience of onboard systems and equipment
Who it is aimed atThe ship, through the systems integrator, shipowner and class societyThird party equipment suppliers
Structure17 requirements across identify, protect, detect, respond and recover30 security capabilities for all systems in scope, plus 11 more for systems connected to untrusted networks, based on IEC 62443-3-3
Key technical pointsVessel asset inventory of hardware and software kept up to date for the ship's life; security zones and firewalled boundaries; remote access rules; mobile and portable device control; network monitoring; incident response and recoveryIdentification and authentication, malicious code protection, least functionality, audit records, backup and restore, secure remote access capabilities
Development processCyber resilience test procedure for construction, commissioning and annual surveysSecure development lifecycle based on IEC 62443-4-1: protect code signing keys, document security updates, deliver authentic updates, test updates before release
ScopePropulsion, steering, anchoring and mooring, power, fire, bilge and ballast, loading computer, watertight integrity, lighting, required safety systems, navigation and communication systems, and any IP interface from these to other systemsSystems and equipment that fall in E26 scope

Sources: IACS UR E26 Rev.1, IACS UR E27 Rev.1, ABS, 2023.

For navigation and radiocommunication systems, E26 lets class accept IEC 61162-460 or equivalent standards in place of E27's security capabilities, provided E26 itself is met and those systems sit in their own zone.

What this means in practice

Three points are worth drawing out.

E26 asks for a software inventory. The vessel asset inventory must cover "application programs, operating systems, if any, firmware and other software components," must be kept up to date "during the entire life of the ship," and must record software changes that could introduce new vulnerabilities. E27 tells suppliers what to provide: for each software component, its maker, model, purpose and version. That is close to a software bill of materials (SBOM), a machine readable list of every component in a piece of software, and SBOMs are the practical way to meet it at scale.

E27 is a supply chain rule. It reaches past the shipowner to the equipment maker's development process, signing keys and update documentation. Suppliers must tell users which product versions a security patch applies to, how to apply and verify it, and what the risk is of not applying it, and must document security updates for dependent components and operating systems.

Most of the world fleet sits outside E26 and E27. They apply to new contracts from July 2024. Existing ships rely on MSC.428(98), flag and class requirements, and industry guidance such as the BIMCO guidelines, which reached version 5 in 2024 and are produced and supported by a long list of shipping associations, class societies, insurers and companies.

For a deeper look at the ISM Code side, see our post on maritime vessel OT cyber security and the ISM Code. For the IEC 62443 family that E27 builds on, see IEC 62443-2-2 security protection schemes.

Why is maritime security a software supply chain problem?

Because almost all the software on a ship or in a terminal is written by someone else. Component makers and open source projects feed equipment makers, who feed integrators and shipyards, who hand ships to owners. Updates flow back the other way through vendor portals, remote sessions and USB media. Most documented failures trace to a link in that chain.

Diagram of the maritime software supply chain from components to equipment makers, integrators and yards, owners and managers, and the vessel, with updates flowing out to the ship, the rules that cover each link (UR E27, UR E26, MSC.428(98)) and documented failure points

Figure 10. Who builds the software on a ship, how updates reach it, and which rule covers each link.

Look back at the documented cases through that lens:

  • NotPetya (2017). Delivered through a hijacked update server for third party accounting software, then spread across a shipping line's network.
  • Satcom CVEs (2014 to 2026). Hard coded and default credentials, and authentication bypasses, built into firmware by equipment makers and inherited by every ship that installed it.
  • Intellian firmware (2018). Firmware for maritime antenna controllers was downloadable from openly readable cloud storage, which is how a researcher found hard coded credentials in it.
  • AmosConnect 8 (2017). A shipboard product with a backdoor account and no fix, because it was already end of life.
  • DNV ShipManager (2023). Ransomware on a fleet software provider's servers reached about 1,000 ships' shore side functions.
  • BIMCO's failed ECDIS update. A manufacturer's update that the ship's old operating system could not run took down every ECDIS at sea.
  • The worm in the power management system. Brought aboard on a service technician's USB device.

Shipowners and terminal operators rarely see inside the software they buy. An ECDIS is a Windows application with chart engines and libraries. A satcom terminal is embedded Linux with a web server, BusyBox, OpenSSL and the maker's own code. A TOS is a large enterprise application on databases and middleware. Each of those components has its own vulnerabilities, published over time, and the owner usually cannot tell whether a new CVE affects their ships until the vendor says so.

That is why E26's software inventory and E27's secure development lifecycle matter more than any single control. They push the industry toward knowing what is inside every system and fixing it on a known timeline. SBOMs are how makers can answer "do our products contain this component, and which version" in minutes instead of weeks. Our guides to SBOMs for firmware and IoT devices, SBOMs for embedded systems and SBOM requirements for industrial control systems go deeper, and VEX (Vulnerability Exploitability eXchange) is how a maker tells customers that a listed vulnerability does not affect a product.

Why it is hard at sea

Ships make the supply chain problem harder in specific ways. Equipment stays in service far longer than its software support. Updates need the maker, sometimes the class society, and a port stay with the right technician aboard. Bandwidth was historically scarce, so updates arrived on USB. And responsibility is split between the owner, the manager, the charterer, the yard and dozens of suppliers. The fix is not one product. It is every party in the chain knowing what it ships and being able to fix it.

How should shipowners and ship managers secure their fleets?

Shipowners should start with the cheap, common attacks: inventory every system and its software, firewall zones and the ship to shore link, control vendor remote access and USB media, change default passwords, harden the fleet domain, keep isolated backups, and train crews to navigate and operate when screens fail.

Three column hardening checklist graphic for shipowners and managers, ports and terminals, and equipment makers

Figure 11. Hardening checklists for the people who own, operate and build maritime systems.

Know what you have

  • Keep a hardware and software inventory for each ship, including firmware and software versions, and update it after every service visit. On new ships, E26 requires this; on existing ships it is still the foundation.
  • Ask equipment makers for SBOMs and for their security advisories, and record support end dates.
  • Map every connection between zones and to shore, including remote access paths.

Separate and protect

  • Group systems into security zones: navigation and comms, machinery, cargo, business IT, crew and guest. Firewall the boundaries and allow only documented traffic.
  • Remove dual-homed PCs and unofficial Wi-Fi access points.
  • Never expose satcom terminal or OT management interfaces to the internet. Change every default password at installation.
  • Put crew and guest networks in their own zone with no route to OT.

Control people and media

  • Route all vendor remote access through one gateway with per vendor accounts, multifactor authentication, onboard approval for each session, and logging.
  • Control USB and removable media, including visitors' drives. Scan media on a dedicated station before it touches bridge or engine computers.
  • Train crews on phishing and on reporting unexplained computer faults.

Harden the fleet domain

  • Tiered administration, few domain administrators, unique local administrator passwords, and multifactor authentication on privileged accounts.
  • Segmentation and firewalls between vessels so one ship's compromise does not reach the next.
  • Endpoint detection on shore and ship Windows machines.

Prepare to operate degraded

  • Keep local backups on each ship and shore backups isolated from the domain. Test restores.
  • Drill navigating without GNSS and without ECDIS, and operating without the loading computer and office systems.
  • Keep paper copies of critical certificates and records where that is cheap.
  • Address all of this in your safety management system, as MSC.428(98) requires, and follow the BIMCO guidelines for detail.

How should ports and terminal operators secure their systems?

Ports should protect the terminal operating system, its database and identity systems as their most critical assets, segment crane, gate and reefer OT from IT, control vendor remote access, keep offline backups and manual fallbacks, and rehearse a multi day outage. Every documented port incident ended with people working by hand.

  • Protect the core. Segment the TOS, its database servers and identity infrastructure. Apply multifactor authentication to all administrative and remote access. Patch internet facing systems quickly.
  • Separate OT. Keep quay crane, yard equipment, gate automation and reefer power networks in their own zones, with only documented traffic to the TOS.
  • Control third parties. TOS vendors, crane makers and integrators need remote access. Give them individual accounts, approve sessions, and log them. Monitor EDI and portal accounts used by shipping lines and hauliers.
  • Back up and rehearse. Keep offline, tested backups of the TOS and its database. Write and rehearse a manual gate and yard procedure for a two day outage, and decide in advance when you would disconnect from the internet, as DP World Australia did.
  • Know your software. Ask TOS and equipment vendors for SBOMs and advisories so you can tell quickly whether a new vulnerability affects you.
  • Share quickly. Port incidents affect shipping lines, hauliers and customs. Agree in advance how you will tell them and what indicators you will share.

How should maritime equipment makers build secure products?

Equipment makers should build to IACS UR E27 and IEC 62443, remove every hard coded and default credential, sign firmware and protect signing keys, scan their own code and the components they ship, publish SBOMs and advisories, test updates before release, and support products for as long as ships will run them.

The satcom CVE table above is a list of things makers can stop doing. Concretely:

Design

  • Build to UR E27's security capabilities and secure development lifecycle, which are based on IEC 62443-3-3 and 62443-4-1.
  • No hard coded credentials and no shared default passwords. Force a unique password at installation.
  • Make remote access off by default, require onboard approval and multifactor authentication, and log every session.
  • Never authenticate in client side code. Treat every diagnostic feature that runs a command as an attack surface.

Build and release

  • Scan your own code with static analysis and secrets scanning, and scan every open source and third party component in each release.
  • Scan the binaries and images you actually ship, not just the source, because that is what ends up on ships.
  • Produce an SBOM for every release and give it to integrators and owners. UR E26 asks them for software inventories; your SBOM is how they get one.
  • Sign firmware and updates, protect signing keys as UR E27 section 5.1 requires, and add rollback protection.
  • Test updates against the operating systems and configurations in the field before release, so you do not repeat BIMCO's ECDIS crash.
  • Keep firmware download locations access controlled.

Support

  • Publish security advisories and patch documentation with the details UR E27 asks for: affected versions, how to apply and verify, impact, and the risk of not patching.
  • State support periods in contracts, and do not abandon products that ships will run for years.
  • Run a coordinated disclosure process and respond to researchers. Several advisories cited here exist because researchers reported flaws and makers fixed them.

Our guides to firmware supply chain security and OT software supply chains in manufacturing cover these practices in more depth.

How Safeguard helps teams that build maritime software

Safeguard covers the software supply chain inside maritime products in one platform: SBOMs for firmware, apps and terminal software, binary and image scanning, vulnerabilities ranked by reachability, SCA, SAST and secrets scanning on code you own, fix pull requests gated by your CI, Gold patched packages, supplier risk and compliance evidence.

Safeguard does not sit on a ship's NMEA bus, listen to AIS or watch the bridge network. It sits where most of the long lived exposure in this guide comes from: the firmware, applications, open source components and services that equipment makers, integrators, fleet software providers and terminal system vendors build and maintain. Teams that take this seriously often end up with one tool for source scanning, another for binary analysis, another for SBOM management, a questionnaire tool for suppliers and a GRC platform for audits. Safeguard does that work end to end, with one inventory and one vulnerability model, from the first SBOM to the merged fix and the audit record.

Flow diagram of Safeguard for maritime software: source repositories, binaries and images, extracted firmware files and supplier SBOMs feed Safeguard, which produces fix pull requests, Gold patched packages, supplier risk scoring, compliance evidence and SBOMs for shipyards and owners, deployable as SaaS, VPC or on premises

Figure 12. How Safeguard fits maritime software teams, from inputs to fixes and evidence.

See inside the binaries and images you ship and receive

In the Safeguard console, Integrations > Upload App Binary accepts compiled software: Windows EXE, MSI and DLL files, Android APKs and iOS IPAs, VM and disk images in QCOW2, VMDK, VHD and OVA formats, and container tarballs. Safeguard unpacks the file, identifies embedded libraries and their versions, maps them to known vulnerabilities and records licences, and the result becomes a project with a full SBOM. For maritime teams that covers Windows based ECDIS, loading computer and engine HMI software, crew and fleet apps on both mobile platforms, virtual appliances used for terminal and fleet systems, and containerised backends. Raw firmware images for satcom terminals, controllers and embedded bridge devices may need to be extracted into their component files first, and the usual limits of binary analysis apply: heavy obfuscation, encryption or unusual packing can give incomplete results. Our guide to scanning app binaries with Safeguard and our explainer on binary composition analysis go further.

Shipowners and terminal operators can use the same feature on what they receive: upload a vendor's installer or appliance image and see what is inside before it goes aboard or into the terminal network.

Rank vulnerabilities by what can actually be reached

A satcom terminal image or a TOS release can contain hundreds of components, and many vulnerabilities in them sit in code the product never calls. Safeguard's reachability analysis ranks the vulnerabilities your code can actually reach above those it cannot, and adds EPSS (the Exploit Prediction Scoring System) and known exploited context. That is how a maker tells the web interface flaw that needs fixing this week from an unused function in a bundled library, and it gives a UR E27 security update decision the evidence it needs.

Scan the code you own

For bridge applications, OT software, terminal systems, crew and fleet apps and cloud services that you write yourself, Safeguard runs software composition analysis (SCA) across the full transitive dependency tree, static application security testing (SAST) on your own code, and secrets scanning that catches hard coded passwords, API keys and signing keys before they ship. Hard coded credentials (CVE-2014-2940, CVE-2020-8000), client side only authentication (CVE-2018-16590) and command injection through diagnostic fields (CVE-2019-17269) are exactly the classes that static analysis and secrets scanning exist to catch.

SBOMs for shipyards, owners and class

Every project produces an SBOM you can export in CycloneDX or SPDX, among other formats. For an equipment maker, that is the software component list UR E27 asks suppliers to provide and the raw material for the vessel asset inventory that UR E26 asks integrators and owners to keep. For an owner or terminal operator, Safeguard's Third-Party Risk Manager lets you request SBOMs from suppliers, score them and monitor them continuously, so the supplier side of the chain is visible too. See our post on third-party risk management with Safeguard.

Fix it without waiting for upstream

This is where Safeguard differs most from a scanner. For code in your repositories, Griffin, Safeguard's remediation model, opens a fix pull request that your own CI builds and tests before anyone merges it, and your policies decide what can merge. For open source components where upstream has not shipped a fix, or where the fix sits in a version you cannot move to without a larger change, Safeguard Gold provides patched, hardened builds of those packages, and Safeguard can fork and patch the vulnerable package so your next build pulls a fixed version. For an equipment maker waiting on an upstream library, or an integrator waiting on a maker, that can be the difference between a fix in the next scheduled update and a fix next year. Closed source components still have to come from their vendor, but the open source around them no longer has to wait.

Compliance evidence from the same data

Safeguard's compliance suite collects evidence for frameworks such as SOC 2 and ISO 27001 from the same scans, policies and fix history. For teams working to IACS UR E26 and E27, IEC 62443 or a shipowner's MSC.428(98) program, the SBOMs, vulnerability monitoring records, risk decisions and fix trail are the raw material that class surveys, customer reviews and audits ask for.

Deploy where your code lives

Bridge software source, OT code and signing infrastructure are sensitive. Safeguard runs as SaaS, in your own VPC, or fully on premises. It can run in Zero AI mode with no language model in the loop, or with a model you choose.

How this compares

Maritime teams often combine several tools: a maritime security platform such as Cydome, which describes "built-for-maritime cybersecurity that protects your IT and OT in real-time"; OT network monitoring from vendors such as Nozomi Networks; a source SCA tool such as Snyk or Black Duck; and a GRC tool for audits. Those are capable products, and onboard and terminal network monitoring does a job Safeguard does not: watching live traffic on ships and in ports. Safeguard's case is consolidation and remediation across the software itself: one inventory across source, binaries and images; one vulnerability model with reachability; fix pull requests and patched packages instead of tickets; and supplier risk and compliance evidence from the same data. Safeguard can work alongside an existing maritime or OT monitoring platform and take over SBOM delivery and fixing.

The takeaway

Ship hacking has a cinematic image: a stranger on a laptop steering a container ship into a bridge. The record looks different. A trusted software update that wiped a shipping line's network. A fleet software provider's servers encrypted. A port's terminal system down for two days. Satellite terminals with passwords printed in the manual. A worm on a service technician's USB stick. And around all of it, AIS and GNSS signals that anyone in range can fake or drown.

The fixes share a shape. Know exactly what software is on every ship and in every terminal. Separate the zones and control who reaches in. Keep the crew able to navigate and operate when the screens lie or go dark. And get the software supply chain to do its part: makers shipping without hard coded credentials, with SBOMs and signed updates, and owners able to fix what they run without waiting for everyone upstream. IACS UR E26 and E27 push in exactly that direction for new ships. The rest of the fleet, and every port, does not need to wait for a rule to do the same.

Try it yourself

Click through these interactive Safeguard demos to see the steps described in this guide.

Test yourself

Where do most documented cargo ship and port attacks happen?

The documented record is mostly shore side. Ransomware and destructive malware have stopped shipping lines, ship managers and port terminals, and no publicly reported ransomware incident has been confirmed to have run on a vessel's own computers.

Why does AIS spoofing work?

AIS has no message authentication. Receivers accept any well formed message on the AIS channels, and the identity number inside it is a claim, not a proof.

What is the difference between GNSS jamming and GNSS spoofing?

Jamming drowns satellite navigation signals so receivers lose their position, and it usually triggers alarms. Spoofing replaces the real signals with counterfeit ones so receivers report a wrong but plausible position, often with no alarm. Cross checking against radar, gyro and log counters both.

What happened to Maersk in June 2017?

A destructive malware outbreak later called NotPetya spread from a compromised M.E.Doc software update. Maersk said it mainly hit Maersk Line, APM Terminals and Damco, and expected a negative impact of USD 200 to 300 million.

Why is vendor remote access such a large risk, and what does IACS UR E26 require?

It is a trusted, privileged path that bypasses most controls, and installers leave remote tools running or share accounts. UR E26 sets rules including onboard approval, multifactor login, encrypted tunnels, session limits and logging.

What is a safe order for fixing maritime cyber risks?

Fix the cheap and damaging ones first: ransomware through shore and fleet IT, exposed satcom terminals, unmanaged vendor access and infected USB media. Then plan and monitor for harder cases like targeted GNSS spoofing, NMEA tampering and supplier compromise.

Frequently asked questions

Can a cargo ship be hacked?

Yes. Ships run many networked computers, and researchers have shown that satellite terminals, ECDIS units and NMEA links can be compromised, and that AIS and GNSS can be spoofed. Real incidents so far have mostly hit shore systems of shipping companies and ports through ransomware and malware. Pen Test Partners says no publicly reported ransomware incident has been confirmed on a vessel's own computers, but warns that fleet-wide networks make it increasingly possible.

What is AIS spoofing?

AIS spoofing is broadcasting false Automatic Identification System messages, such as a ship that does not exist, a real ship in the wrong place, or a fake distress beacon. It works because AIS has no message authentication: receivers accept any well formed message. Trend Micro researchers demonstrated spoofing, hijacking and availability attacks in 2014. Ships detect it by cross checking AIS targets against radar and visual lookout.

What is the difference between GPS jamming and GPS spoofing at sea?

Jamming drowns satellite signals so the receiver loses its position, which usually triggers alarms. Spoofing sends counterfeit signals so the receiver reports a wrong position that looks valid, often with no alarm. In 2013 a UT Austin team spoofed a yacht onto a parallel track with the owner's permission. In March 2026 Windward reported jamming affecting more than 1,100 ships in a day in the Middle East Gulf.

Can hackers take control of a ship's steering?

In research settings, partly. Pen Test Partners showed in 2018 that ECDIS position data could be manipulated and that NMEA messages, which can carry autopilot commands, have no authentication, and the UT Austin team steered a yacht off course by spoofing GPS. Doing this to a real ship needs deep access and an inattentive crew. Radar, visual watchkeeping and manual steering remain the safeguards.

What happened to Maersk in the 2017 NotPetya incident?

In June 2017 destructive malware delivered through a hijacked update for third party accounting software spread across Maersk's network. Maersk said it mainly affected Maersk Line, APM Terminals and Damco and expected a USD 200 to 300 million hit to results. Its chairman later said the company reinstalled 4,000 servers, 45,000 PCs and 2,500 applications in about ten days and handled much of its volume manually.

What do IACS UR E26 and E27 require?

UR E26 sets cyber resilience requirements for the whole ship: an asset inventory including software versions, security zones, remote access controls, network monitoring, and incident response and recovery. UR E27 sets security capabilities and a secure development lifecycle for equipment suppliers, based on IEC 62443. Both apply to ships contracted for construction on or after 1 July 2024.

What does IMO resolution MSC.428(98) require?

MSC.428(98), adopted in June 2017, requires that a ship operator's approved safety management system takes cyber risk management into account under the ISM Code. Flag administrations were encouraged to check this no later than the first annual verification of the company's Document of Compliance after 1 January 2021. It is a management requirement, supported by the guidelines in MSC-FAL.1/Circ.3.

How should a shipowner control vendor remote access?

Route all vendor access through one controlled gateway, give each vendor its own account with multifactor authentication, require a responsible person on board to approve each session, limit sessions to the system being serviced, and log everything. Remove always-on remote tools and never expose system addresses to the internet. These match the remote access rules in IACS UR E26.

Sources

Every factual claim in this guide links to its source inline. The full list, with the exact URLs used, is in the front matter above. Key references:

Never miss an update

Weekly insights on software supply chain security, delivered to your inbox.

We email you a link to confirm, and only then add you to the list. Unsubscribe in one click. Privacy

Self-healing security runs on Safeguard.

Your first fix PR is minutes away.

No sales call required, even your agent can complete the purchase over MCP.