Civil Law and Autonomous Port Infrastructure Failure Litigation in Europe

1. Introduction

Autonomous port infrastructure failure litigation concerns civil, commercial, environmental, and public-law disputes arising when automated or AI-enabled port infrastructure malfunctions and causes damage.

Modern ports increasingly use:

autonomous cranes and cargo-handling equipment;

automated guided vehicles (AGVs);

autonomous trucks and yard vehicles;

automated gates and access systems;

AI-based traffic-management systems;

automated mooring and docking systems;

smart warehouses;

digital-twin infrastructure;

predictive-maintenance systems;

automated navigation and vessel-berthing systems;

robotic inspection systems;

sensor-controlled bridges, locks and barriers;

cybersecurity-controlled infrastructure.

A failure can therefore be partly physical, partly software-based, and partly organisational.

For example, an autonomous crane may receive an incorrect sensor signal, software may incorrectly classify the position of a container, the crane may move automatically, and the resulting collision may damage a ship, container, terminal equipment and cargo. The litigation then becomes a question of who legally caused the failure: manufacturer, software developer, port operator, maintenance contractor, infrastructure owner, system integrator, cybersecurity provider, or public authority.

There is no single European case dealing with the exact combination of AI + autonomous port + infrastructure collapse. The legal analysis therefore combines EU product liability, infrastructure/environmental law, contract law, tort/delict principles, maritime law, and public-authority liability. The cases below are consequently direct or closely analogous authorities, rather than cases all involving autonomous ports.

2. Meaning of Autonomous Port Infrastructure

An autonomous port is a port in which important operational functions are performed partly or substantially through automated systems.

Examples

The important legal feature is that the human decision-maker may not have directly caused the event.

3. European Legal Framework

A. National civil law

The primary civil claim will usually arise under the law applicable to the port and the relevant contract.

Possible causes include:

breach of contract;

negligent maintenance;

defective construction;

defective product;

professional negligence;

breach of safety duties;

failure to warn;

failure to update software;

cybersecurity negligence;

environmental damage;

property damage;

personal injury;

economic loss.

Civil-law systems generally require analysis of:

Duty → Breach → Causation → Damage → Remedy.

4. EU Product Liability

A particularly important development is Directive (EU) 2024/2853 on liability for defective products.

The new Directive expressly treats software, including AI systems, as products for product-liability purposes. It also recognises that software may remain under a manufacturer's control through updates, upgrades and machine-learning processes. The Directive applies to products placed on the market or put into service after 8 December 2026. (Eur-Lex)

This is highly relevant to autonomous ports.

Suppose:

An autonomous crane operates correctly when installed, but a later software update causes its positioning algorithm to malfunction.

Under the new European framework, the fact that the defect appeared after the original installation does not necessarily end the manufacturer's responsibility where the relevant software or related service remained under its control. (Eur-Lex)

5. Critical-Infrastructure Resilience

Ports also have an important public-infrastructure dimension.

The Critical Entities Resilience Directive (EU) 2022/2557 establishes a European framework requiring critical entities to improve their ability to prevent, withstand, respond to, and recover from disruptive incidents. Transport is within the framework, and the Directive expressly recognises interdependencies between infrastructure sectors. (Eur-Lex)

This matters because a port failure may simultaneously affect:

shipping;

road transport;

rail;

electricity;

telecommunications;

customs;

logistics;

fuel supply;

food supply.

Therefore, a court examining negligence may consider not merely whether an individual machine was defective, but whether the operator maintained an adequate system of resilience and risk management.

6. AI Act and Autonomous Infrastructure

The EU AI Act, Regulation (EU) 2024/1689, is also relevant where AI performs safety-critical infrastructure functions.

The AI Act recognises that certain AI systems used as safety components in critical infrastructure can be high-risk because malfunction can threaten health, safety and the physical integrity of infrastructure. (Eur-Lex)

Thus, where an autonomous port system uses AI for safety-related functions, the operator may face overlapping:

AI regulatory obligations;

product-safety obligations;

cybersecurity obligations;

contractual duties;

tort/delict duties;

maritime obligations.

Importantly, regulatory compliance does not automatically eliminate civil liability.

7. Essential Elements of an Autonomous Port Failure Claim

7.1 Duty of care

The claimant must identify the relevant legal duty.

Possible defendants include:

port authority;

terminal operator;

crane manufacturer;

software developer;

system integrator;

maintenance contractor;

cybersecurity provider;

engineering consultant;

vessel owner;

classification organisation;

public authority.

7.2 Breach

Breach may arise from:

defective design;

inadequate testing;

insufficient redundancy;

defective sensors;

inadequate maintenance;

failure to install safety barriers;

failure to patch software;

inadequate cybersecurity;

insufficient human supervision;

failure to investigate warning signals;

failure to conduct appropriate risk assessments.

7.3 Causation

Autonomous systems create particularly difficult causation questions.

A typical chain could be:

Sensor defect → incorrect AI input → incorrect algorithmic decision → autonomous movement → collision → infrastructure damage → business interruption.

The claimant must establish the legally relevant causal connection.

8. Case Law

Case 1 — Boston Scientific Medizintechnik GmbH v AOK Sachsen-Anhalt

CJEU, Joined Cases C-503/13 and C-504/13, 2015

Facts

The case concerned defective medical devices and the risk of failure in products belonging to the same series.

Principle

The CJEU interpreted the EU defective-product regime broadly where products presented an abnormal safety risk. It also accepted that costs associated with eliminating the risk could fall within the relevant personal-injury consequences.

(Eur-Lex)

Relevance to autonomous ports

The reasoning is useful where an autonomous port operator discovers that an entire series of:

autonomous cranes;

robotic vehicles;

sensors;

control units;

contains the same safety defect.

The claimant may argue that the relevant question is not merely whether this particular machine has already caused an accident, but whether the product has an abnormal safety risk.

Legal lesson

Systemic safety defects can be legally important even before catastrophic failure occurs.

9. Case 2 — W and Others v Sanofi Pasteur

CJEU, Case C-621/15, 2017

Principle

The CJEU addressed evidentiary difficulties in product-liability cases where scientific evidence does not establish causation with certainty.

The case is important because the Court recognised that, subject to national evidentiary rules, serious, precise and consistent evidence may contribute to establishing defect and causation.

Relevance

Autonomous infrastructure frequently creates an evidentiary problem:

The system's algorithm is proprietary, continuously learning, and difficult for the claimant to understand.

A claimant may therefore rely on:

incident logs;

repeated failures;

maintenance records;

system warnings;

expert reconstruction;

sensor records;

software-version history;

abnormal operational patterns.

Legal lesson

Complex technology does not automatically make causation legally impossible.

10. Case 3 — Commission v United Kingdom

CJEU, Case C-300/95, 1997

Principle

The case concerned the development-risk defence under the European product-liability framework.

The Court examined what could reasonably be known from the state of scientific and technical knowledge.

Relevance to autonomous ports

Autonomous infrastructure is constantly evolving.

A manufacturer might argue:

“The failure could not have been discovered using the scientific and technical knowledge available when the system was supplied.”

The claimant may respond that the relevant technology was sufficiently mature that the risk should have been identified.

Example

If an autonomous mooring algorithm had a known failure mode in heavy winds, the manufacturer may have difficulty relying on technological uncertainty if appropriate testing could reasonably have identified it.

Legal lesson

Technological novelty does not automatically equal legal immunity.

11. Case 4 — Commune de Mesquer v Total France

CJEU, Case C-188/07, 2008

This case arose from the Erika oil-tanker disaster.

The sinking caused substantial pollution along the French coast. The CJEU examined the application of European environmental law and the polluter-pays principle.

Principle

The case demonstrated how European environmental law can impose financial consequences for environmental consequences arising from maritime accidents.

Relevance to autonomous ports

Suppose autonomous infrastructure fails and causes:

fuel leakage;

chemical contamination;

destruction of marine habitats;

release of hazardous cargo;

sediment pollution.

The dispute may extend beyond ordinary property damage into environmental liability.

Legal lesson

Port infrastructure failure may create both private economic claims and environmental-remediation obligations.

12. Case 5 — Raffinerie Mediterranee (ERG)

CJEU, Joined Cases C-379/08 and C-380/08, 2010

The cases concerned environmental liability and the polluter-pays principle under the Environmental Liability Directive. (Infocuria)

Principle

The Court considered the relationship between environmental damage, remedial measures and the identification of responsible operators.

Relevance

Consider an autonomous port where a software failure causes:

automated pumps → incorrect chemical transfer → tank overflow → marine contamination.

The operator may face environmental-remediation obligations even if the immediate physical accident resulted from an automated system.

Legal lesson

Autonomous operation does not eliminate the responsibility of the economic operator controlling the activity.

13. Case 6 — Kraaijeveld and Others

CJEU, Case C-72/95, 1996

This case concerned environmental impact assessment for infrastructure works, including flood-relief and dyke-related works.

The CJEU held that Member States' discretion concerning which projects require environmental assessment is subject to the obligation to assess projects likely to have significant environmental effects. (Infocuria)

Relevance to ports

Major autonomous port infrastructure projects may involve:

new automated terminals;

enlarged breakwaters;

autonomous shipping corridors;

automated container yards;

dredging;

new transport connections.

If authorities inadequately assess environmental consequences, later litigation may challenge the authorisation or seek remedies under applicable national and EU law.

Legal lesson

Infrastructure autonomy does not remove environmental assessment obligations.

14. Case 7 — Gemeinde Altrip and Others

CJEU, Case C-72/12, 2013

The case concerned challenges to development-consent decisions and environmental impact assessment procedures. (Infocuria)

Relevance to autonomous ports

Suppose a port authority approves an autonomous terminal without properly considering:

collision risks;

environmental effects;

traffic changes;

emergency scenarios;

cumulative infrastructure effects.

Affected persons may challenge the legality of the underlying authorisation, depending on the applicable national procedural system.

Legal lesson

Procedural environmental defects can become legally significant when infrastructure causes later damage.

15. Case 8 — Öneryıldız v Turkey

ECtHR, Grand Chamber, 2004

This is an important European public-authority liability analogy.

The case concerned an explosion at a municipal rubbish tip that caused deaths and destruction of property. The European Court of Human Rights found violations relating to the authorities' obligations concerning dangerous activities and effective protection of life and property. (HUDOC)

Relevance to autonomous ports

A port authority may operate infrastructure involving inherently dangerous activities:

cranes;

fuel terminals;

hazardous chemicals;

heavy machinery;

autonomous vehicles;

high-voltage systems.

Where authorities know about a serious infrastructure risk but fail to take reasonable preventive measures, public-law liability may arise alongside ordinary civil claims.

Legal lesson

Authorities responsible for dangerous infrastructure cannot necessarily rely on the fact that an accident was technically caused by an automated system.

16. Case 9 — Budayeva and Others v Russia

ECtHR, 2008

The case concerned a dangerous natural event and a State-controlled protective structure. A damaged mud-retention dam had not been adequately maintained; warnings were not effectively implemented, and the subsequent disaster caused deaths, injuries and property destruction. (HUDOC)

Relevance to autonomous ports

This reasoning is particularly useful for maintenance failures.

Imagine:

An autonomous port's structural-monitoring system repeatedly reports abnormal corrosion in a quay wall.

The operator ignores the warnings.

The quay later collapses.

The defence cannot necessarily be:

“The AI failed.”

A court may instead ask:

Who received the warning?

Was human intervention required?

Was maintenance overdue?

Was the risk foreseeable?

Was the monitoring system appropriately configured?

Were emergency procedures implemented?

Legal lesson

Failure to respond to known infrastructure risks can be more important than the fact that the system was autonomous.

17. Case 10 — Erika / Total

French Cour de cassation, Criminal Chamber, 25 September 2012

The Erika litigation is particularly relevant to maritime infrastructure and environmental loss.

The French Court of Cassation confirmed significant civil consequences arising from the oil spill and recognised compensation relating to ecological damage. The French Ministry of Justice describes the judgment as establishing the principle of ecological damage in French jurisprudence. (Ministère de la justice)

Relevance to autonomous port failures

Suppose an autonomous port system causes:

oil contamination;

destruction of marine ecosystems;

coastal pollution;

fisheries losses;

tourism losses.

The resulting litigation may include both traditional economic loss and ecological damage.

Legal lesson

Environmental loss can become an independent head of civil compensation under national law.

18. Case 11 — Prestige Litigation

The Prestige oil-spill litigation illustrates the complexity of allocating civil liability among maritime actors.

Spanish proceedings ultimately imposed civil liability on multiple parties, with the Spanish Supreme Court's 2018 judgment concerning the master, owners and the P&I insurer and subsequent quantification of claims. (Eur-Lex)

Relevance

An autonomous port accident may similarly involve a chain of actors:

software provider → equipment manufacturer → integrator → port operator → vessel → cargo owner → maintenance company.

The Prestige litigation demonstrates why maritime accidents frequently cannot be reduced to one defendant.

19. Autonomous Port Failure: Main Liability Models

A. Manufacturer liability

A manufacturer may be liable where:

hardware is defective;

software is defective;

sensors are unreliable;

safety systems are inadequate;

warnings are insufficient;

design does not provide appropriate safeguards.

Under the new Product Liability Directive, software and AI systems are expressly brought within the product-liability concept. (Eur-Lex)

20. Software Developer Liability

Software may cause infrastructure failure without any physical component breaking.

Examples:

incorrect path planning;

defective object recognition;

wrong container identification;

erroneous load calculations;

faulty collision avoidance;

defective machine-learning model;

inadequate software update.

The new EU product-liability framework is especially significant because it treats software as a product and addresses defects arising through software updates and upgrades under the manufacturer's control. (Eur-Lex)

21. Port Operator Liability

The port operator may be liable for:

Poor supervision

Even an autonomous system may require human oversight.

Poor maintenance

A machine can be technologically sophisticated but physically deteriorated.

Failure to respond to warnings

Repeated alarms may establish knowledge of risk.

Poor emergency planning

The operator should consider:

loss of connectivity;

sensor failure;

GPS failure;

cyberattack;

power failure;

extreme weather;

algorithmic malfunction.

Excessive reliance on automation

A port cannot necessarily transfer all operational responsibility to an AI system.

22. System Integrator Liability

The system integrator is particularly important.

An individual component may be safe by itself.

However:

Sensor + AI + crane + network + control software

may become unsafe when integrated.

Therefore, the integrator may face liability if it:

selected incompatible components;

failed to test interoperability;

improperly configured software;

failed to establish safe fallback mechanisms;

ignored known integration risks.

23. Cybersecurity Failure

Autonomous ports create an important cybersecurity dimension.

Imagine:

Cyberattack → manipulated sensor data → AI receives false information → autonomous crane moves incorrectly → ship collision.

Possible defendants may include:

cybersecurity contractor;

software developer;

port operator;

infrastructure owner;

network provider.

The critical question becomes whether the relevant actor had a legal duty to maintain an appropriate level of cybersecurity.

The new Product Liability Directive expressly recognises cybersecurity vulnerabilities and software updates as relevant to continuing product safety. (Eur-Lex)

24. Sensor Failure

Autonomous systems depend heavily on sensors.

Examples:

radar;

cameras;

LiDAR;

GPS;

pressure sensors;

load sensors;

vibration sensors;

environmental sensors.

A sensor may give a false reading.

Legal problem

Suppose:

Sensor says “no vessel present” → autonomous crane/vehicle operates → vessel is struck.

The court may need to determine:

Was the sensor defective?

Was its calibration inadequate?

Was redundancy required?

Did the AI properly process the signal?

Should the system have stopped?

Did the operator ignore warning signs?

25. AI Decision-Making and Human Oversight

A central question will be:

Who had the ability and duty to intervene?

Autonomy does not necessarily eliminate human responsibility.

Courts may examine:

supervisory procedures;

intervention thresholds;

emergency stop mechanisms;

staffing;

operator training;

monitoring arrangements;

escalation procedures.

If the system was designed to operate without human intervention, the manufacturer may face stronger arguments concerning the adequacy of its safety architecture.

26. Infrastructure Owner Liability

The owner of the physical infrastructure may face liability for:

structural deterioration;

inadequate inspections;

defective foundations;

inadequate load capacity;

corrosion;

poor electrical infrastructure;

inadequate drainage;

insufficient redundancy.

Autonomous technology does not convert an unsafe physical structure into a safe one.

27. Contractual Liability

Port operations commonly involve multiple contracts.

For example:

Port Authority → Terminal Operator → Automation Provider → Maintenance Company → Software Developer

A failure can therefore produce:

breach-of-contract claims;

indemnity claims;

warranty claims;

limitation-of-liability disputes;

insurance disputes;

contribution claims.

Important contractual clauses include:

Performance warranties

Was the autonomous system contractually guaranteed to achieve a particular safety or performance level?

Service-level agreements

What uptime was promised?

Maintenance obligations

Who was responsible for updates and repairs?

Cybersecurity clauses

Who was responsible for security patches?

Force majeure

Can a cyberattack or extreme weather be treated as force majeure?

Limitation clauses

Are indirect losses or business-interruption losses excluded?

28. Damage Categories

An autonomous port accident can produce several types of damage.

1. Physical damage

quay;

crane;

ship;

containers;

warehouse;

vehicles.

2. Personal injury

workers;

drivers;

passengers;

contractors;

visitors.

3. Cargo damage

Containers may be destroyed or contaminated.

4. Economic loss

A port shutdown may cause:

lost shipping revenue;

delayed deliveries;

contractual penalties;

supply-chain losses.

5. Environmental damage

marine pollution;

habitat destruction;

coastal contamination;

fisheries damage.

6. Business interruption

A major autonomous-system failure could close a terminal for weeks.

29. Causation in Autonomous-Port Litigation

Causation may be the hardest issue.

Traditional accident

Broken crane → falling container → damage.

Autonomous accident

Sensor malfunction → data error → algorithmic interpretation → control command → actuator response → physical movement → collision.

There may therefore be several potential causes.

A court may use:

expert evidence;

software logs;

sensor records;

digital twins;

maintenance records;

system architecture;

audit trails;

CCTV;

cybersecurity records;

testing documentation.

30. Black-Box Problem

AI systems can create an evidentiary problem.

The claimant may know:

“The machine behaved incorrectly.”

But may not know:

“Why did the machine behave incorrectly?”

This creates information asymmetry between:

claimant;

port operator;

software developer;

manufacturer.

The legal significance of this problem depends on the applicable national procedural and evidentiary rules, but European product-liability reforms increasingly recognise the practical difficulty of proving defects in complex technologies.

31. Failure to Update Software

This is an increasingly important issue.

Suppose a manufacturer discovers a cybersecurity vulnerability but does not issue a necessary update.

Later:

Hacker exploits vulnerability → autonomous port equipment malfunctions → accident.

Liability could potentially arise from:

original design defect;

inadequate cybersecurity;

failure to issue an update;

failure to warn;

negligent maintenance.

The 2024 Product Liability Directive specifically addresses defects arising from software and cybersecurity updates. (Eur-Lex)

32. Environmental Liability

Ports are environmentally sensitive locations.

An autonomous failure could cause:

oil spill;

fuel leakage;

chemical release;

underwater pollution;

destruction of marine habitats.

The Environmental Liability Directive and the polluter-pays principle become particularly important.

The ERG litigation illustrates how European environmental liability can involve remedial measures and allocation of responsibility for environmental damage. (Infocuria)

33. Public Authority Liability

A port may be:

privately owned;

publicly owned;

operated through concession;

jointly controlled.

If a public authority:

failed to inspect;

ignored known structural defects;

granted an unsafe authorisation;

failed to enforce safety requirements;

public-law liability may become relevant.

The Francovich/Brasserie du Pêcheur line of CJEU jurisprudence establishes the principle that Member States can be required to compensate individuals for sufficiently serious breaches of EU law where the required conditions, including causation, are satisfied. (Eur-Lex)

This is particularly important where the port failure is connected with a regulatory failure rather than simply private negligence.

34. Francovich and Brasserie du Pêcheur

Francovich principle

A Member State may be liable for damage resulting from breaches of EU law attributable to it.

Brasserie du Pêcheur principle

The CJEU developed the conditions for State liability, including:

the breached rule must confer rights on individuals;

the breach must be sufficiently serious;

there must be a direct causal link between the breach and the damage.

(Eur-Lex)

Application to ports

Suppose EU environmental or infrastructure-resilience requirements are seriously violated by a public authority, and that violation directly contributes to a port catastrophe.

A State-liability claim could potentially arise, subject to the relevant EU rule and national procedural law.

35. Autonomous Port and Critical Infrastructure

The Critical Entities Resilience Directive is particularly significant because it defines resilience broadly as the ability to:

prevent;

protect against;

respond to;

resist;

mitigate;

absorb;

accommodate;

recover from incidents.

(Eur-Lex)

Therefore, modern port litigation should increasingly examine not merely:

“Why did the machine fail?”

but also:

“Why was the overall port system unable to absorb the failure?”

36. Redundancy as a Legal Issue

Autonomous infrastructure should normally be designed around redundancy where the consequences of failure are serious.

Examples:

dual sensors;

emergency manual controls;

backup power;

redundant communication;

alternative navigation systems;

manual override;

emergency shutdown.

Failure to provide reasonable redundancy may support a negligence or defective-design claim where such safeguards were reasonably expected.

37. Force Majeure

Defendants may argue:

extreme storm;

unexpected natural disaster;

cyberattack;

GPS outage;

satellite failure;

power-grid failure.

But force majeure generally depends on the applicable contract and national law.

The central question may be:

Was the event genuinely unforeseeable and unavoidable, or should the operator have designed the system to tolerate it?

For autonomous infrastructure, resilience engineering can therefore affect the legal analysis of force majeure.

38. Contributory Negligence

The defendant may argue that the claimant contributed to the loss.

Examples:

ship entered a restricted autonomous zone;

terminal employee bypassed safety procedures;

cargo was improperly declared;

maintenance instructions were ignored;

emergency warnings were disregarded.

The effect depends on the applicable national law.

39. Multi-Party Liability

A single autonomous port failure may involve:

This makes contribution and allocation of liability especially important.

40. Hypothetical Example

Facts

A European container port operates an autonomous crane.

The crane uses:

AI vision;

LiDAR;

GPS;

cloud software;

automated collision avoidance.

The manufacturer releases a software update.

After the update:

the AI misidentifies a vessel;

the collision-avoidance system fails;

the crane moves into the vessel's path;

a container falls;

the vessel is damaged;

hazardous cargo leaks;

the port closes for ten days.

Potential claims

Vessel owner

May claim:

repair costs;

loss of use;

consequential maritime losses.

Cargo owner

May claim:

destroyed cargo;

delay losses.

Port operator

May claim:

business interruption;

repair expenses;

software-provider indemnification.

Environmental authorities

May seek:

pollution remediation;

restoration costs.

Injured workers

May bring:

personal-injury claims.

41. Possible Defendants

The litigation might proceed against:

Manufacturer

For defective hardware or safety design.

Software provider

For defective update.

AI developer

For defective AI functionality.

Port operator

For inadequate monitoring or human supervision.

Maintenance contractor

For failure to maintain the crane.

Cybersecurity contractor

If a cyber vulnerability contributed to the incident.

Infrastructure owner

If physical infrastructure contributed to the accident.

42. Important Legal Questions for the Court

The court would likely need to determine:

What exactly failed?

Was there a product defect?

Was the software defective?

Was the update responsible?

Was the physical infrastructure adequate?

Was the AI appropriately tested?

Was there adequate human supervision?

Was the risk foreseeable?

Was the system properly maintained?

Did the claimant contribute to the accident?

Was there an intervening cause?

Did environmental damage occur?

Were contractual liability limits applicable?

Which national law governs?

Which defendant legally caused the damage?

43. Special Importance of the 2024 Product Liability Directive

The new Directive represents a significant shift for autonomous infrastructure because it expressly recognises:

software as a product;

AI systems as products;

software supplied through cloud/SaaS arrangements;

defects arising through updates;

cybersecurity-related safety problems;

continued manufacturer control over certain software developments.

(Eur-Lex)

However, it is important to distinguish this new regime from older CJEU cases decided under Directive 85/374/EEC. The new Directive applies to products placed on the market or put into service after 8 December 2026. (Eur-Lex)

44. Case-Law Summary

45. Key Legal Principles

Principle 1

Autonomy does not automatically remove human or corporate responsibility.

Principle 2

Software can be legally relevant as a defective product.

Principle 3

An autonomous system may remain under the manufacturer's responsibility after deployment where relevant software or updates remain under its control.

Principle 4

Port operators can be liable for inadequate maintenance, supervision and emergency planning.

Principle 5

Infrastructure owners can remain responsible for physical defects even where an AI system operates the infrastructure.

Principle 6

Environmental damage can create separate liability from ordinary property damage.

Principle 7

Public authorities can face liability for sufficiently serious breaches of EU law or human-rights obligations, subject to the applicable legal framework.

Principle 8

Causation is likely to be the most technically difficult part of autonomous-port litigation.

Principle 9

Cybersecurity failures can become product-safety and civil-liability issues.

Principle 10

The court will often analyse the entire socio-technical system rather than treating the AI algorithm as an isolated cause.

46. Future Legal Development

The major future issue is the movement from:

“Who operated the machine?”

towards:

“Who designed, trained, integrated, updated, monitored and controlled the autonomous system?”

European liability law is therefore likely to become increasingly concerned with the entire technological chain:

Hardware → Software → AI → Sensors → Network → Human Oversight → Infrastructure → Environment.

For ports, this is particularly important because a failure can create simultaneous contractual, tortious, product-liability, maritime, environmental, cybersecurity and public-authority claims.

47. Exam-Ready Conclusion

Autonomous port infrastructure failure litigation in Europe is a developing area of civil liability involving the interaction of traditional civil-law principles with AI, product liability, environmental law, maritime law and critical-infrastructure regulation. The central issues are defective design, software malfunction, sensor failure, cybersecurity, maintenance, human supervision, causation, environmental damage and allocation of responsibility among multiple actors.

The jurisprudence in Boston Scientific, W v Sanofi Pasteur, Commission v UK, Commune de Mesquer, ERG, Kraaijeveld, Altrip, Öneryıldız, Budayeva, Erika and Prestige provides useful principles concerning defective products, causation, infrastructure safety, environmental damage, public-authority responsibility and maritime liability. The new EU Product Liability Directive 2024/2853, together with the AI Act and Critical Entities Resilience Directive, makes the legal framework increasingly adapted to autonomous infrastructure. (Eur-Lex)

In simple words: when an autonomous port fails, the law does not simply ask “Did the AI make a mistake?” It asks who created the risk, who controlled it, who could have prevented it, whether the risk was foreseeable, and what damage was legally caused by that failure.