Electronic evidence is transforming casualty investigations at sea

What shipboard electronics record

A modern ship carries dozens of sensors and systems that log position, speed, rudder commands, engine settings, radar contacts and collision warnings. When something goes wrong, these records can show what the crew saw, when they acted and what the vessel did in response ^[1].

The Nautical Institute, the professional body for maritime officers, has described this as "a real transformation" of casualty investigations. Electronic records make it possible to reconstruct an incident in a way that paper logs never could ^[1]. The same transformation is happening in other high-consequence settings: autonomous vehicles, medical devices, aviation and financial trading.

The authenticity problem

Electronic records are only useful if people can trust them. A log file that its operator can edit is easy to challenge. The Nautical Institute's briefing on electronic evidence lists the questions a court or arbitrator will ask: Was the data preserved properly? Can you show it hasn't been altered? Is the system that recorded it reliable? ^[1]

These are not rhetorical questions. In disputes over navigation, cargo damage or collisions, opposing parties routinely question the integrity of the other side's records. A screenshot of a radar display or a printout of an engine log has to stand up to that scrutiny.

The law in England and Wales, which governs many maritime disputes, assesses electronic evidence case by case. There is no presumption that a digital record is reliable simply because it came from a computer ^[1]. Someone has to show that the record was kept properly and hasn't been tampered with.

How a digital witness fits shipboard records

A digital witness seals the fingerprint of a record at the moment it is created, without copying the content. That fingerprint goes onto WitniumChain, where any later change would show, even one made by us. When the record is questioned months or years later, anyone you share it with can calculate the fingerprint again and check that it matches ^[4].

For shipboard systems, the thing to seal is each log entry or data file as it is written: the radar plot at 03:42, the engine log for the watch, the autopilot commands during the approach. The data itself stays on the vessel. The fingerprint is what gets sealed.

Who checks it? Surveyors, opposing counsel, arbitrators and courts. They calculate the fingerprint of the record you give them and compare it with the sealed one. If it matches, the record is unchanged since the moment it was sealed. Change a single character and it won't match.

What it shows: that the record existed at that time and hasn't been altered since. What it doesn't show: that the sensor reading was accurate, that the crew acted reasonably or that the incident was unavoidable. It is a record of what was recorded, not proof that everything in it is true.

How it would work

1. The ship's data logging system seals each log file or sensor reading when it is written, sending a fingerprint to WitniumChain through the REST API or SDK.
2. If the vessel is involved in a casualty, the records are handed over as usual, together with the witness IDs.
3. The other party, the surveyor or the court calculates the fingerprint and checks it against the sealed one.

The content never leaves your systems. The first 2,000 witnesses a month are free. Read more at witniumchain.com/docs.

The same question in other sectors

The question the Nautical Institute is asking about ship systems is the one every organisation with automated decision-making will have to answer. The EU's AI Act requires high-risk AI systems to log events automatically, for traceability ^[2]. The logs must be kept for at least six months ^[3]. What the regulation does not say is how to protect those logs from quiet editing.

Chatbots face a simpler version of the same problem. When a customer claims the bot promised something it shouldn't have, the customer's screenshot is often the only evidence. A digital witness per conversation gives the company its own record, sealed when the conversation ends.

In product liability, software now counts as a product under the new EU directive, which covers anything placed on the market after 9 December 2026 ^[5]. If a court wants to know which version of the software was running on a given day, could you show it? A sealed release log makes that a short conversation.

What this means for you

If you operate systems that make decisions, control machinery or talk to customers, ask:

• Could you show what your systems recorded on a specific day six months ago?
• If someone questioned a log, access record or decision trail, what would you hand over?
• Does the evidence you keep have a form that will hold up when it is challenged?

Electronic records are useful until the day they are disputed. Then they need to be provable.

What to watch

For high-risk AI systems under Annex III of the AI Act, the logging requirement in Article 12 applies from 2 December 2027 ^[2]. For Annex I systems, it applies from 2 August 2028. The six-month retention rule applies from the same dates ^[3].

The new Product Liability Directive, which treats software as a product, covers products placed on the market after 9 December 2026 ^[5]. Courts in member states will be able to order defendants to hand over relevant evidence.

For vessels, the requirement is simpler: if you rely on electronic records in a casualty investigation or a dispute, those records have to stand up to scrutiny now.