An impossible AIS track off Batanes, and the difference between what it proves and what it suggests

On 28 September the Philippine Coast Guard said a Chinese-flagged survey vessel had been broadcasting false positions while holding station inside the Philippine exclusive economic zone, near waters crossed by submarine cables ^[1][2]. The vessel is the Jia Hai Ke 7, a 2,000-tonne, 70-metre seabed survey ship ^[1][3]. The evidence offered is a transponder trail that cannot be true: having left Xiamen past midnight on 21 September, its reported position by that evening was off Bindari, Papua New Guinea, more than 2,000 nautical miles away, and it was then detected roughly 38 nautical miles north-west of Itbayat in Batanes ^[1][2][3].

That is a clean finding, and worth separating from everything attached to it.

What the track establishes

A 70-metre survey ship cannot cover 2,000 nautical miles in a day. If the same vessel identifier appears in two seas within that interval, at least one report is false. The PCG called the trail physically impossible, and on the arithmetic it is ^[2][3]. The vessel was also observed loitering rather than transiting, and ignored repeated radio challenges from a patrol aircraft crew asking it to explain its intentions and cease unauthorised activity ^[1][3][4].

Rear Admiral Jay Tarriela, the PCG’s spokesman for the West Philippine Sea, put the concern this way: a seabed-survey vessel hiding its position while holding station in this corridor raises serious concern over the security of the country’s undersea communications infrastructure and the potential military use of the data collected ^[2][4]. The next day the PCG reported challenging a second Chinese vessel off Cabra Island in Occidental Mindoro, describing both as conducting unauthorised activities inside the country’s maritime zone ^[4].

What it does not establish

An independent analysis published this week does the useful work of separating the claims, and it is the most valuable document in this story. It treats the spoofing allegation as the most testable, unauthorised research as a behavioural inference, and the security implication as a risk assessment rather than an observed fact ^[5]. On the first it is firm: if the same 70-metre ship appears in two different seas on the same day, at a distance no such ship could cover, then the data is false ^[5].

On what caused the falsity it declines to choose. The cause could be deliberate manipulation, a transmitter fault, or an error in how a tracking service aggregated the data ^[5]. That is not pedantry. A duplicate or mislabelled feed from an aggregator produces the same artefact as a crew entering a false position, and the published account does not specify which tracking sources recorded each position. Comparing raw tracking records would settle the question either way ^[5]. No such comparison is in the public record.

The analysis is equally direct about the inference chain: a concern about what a survey could enable is not evidence that any cable was surveyed or targeted, and no deployed sampling gear or collected data has been shown ^[5]. Its conclusion is that the strongest element of the account is the physically impossible track, and the weakest link is the chain from loitering to confirmed survey activity to military use ^[5]. R3KONX adopts that reading. The coastal State’s concern is reasonable; the proof published so far supports one link of it.

The legal frame is firmer than the technical one. Under Article 246 of the UN Convention on the Law of the Sea, marine scientific research in the exclusive economic zone and on the continental shelf is conducted with the consent of the coastal State, and Article 248 requires six months’ advance notice with a full description of the project, its methods, vessels and areas ^[9]. Whether research occurred is contested. Whether consent and notification were given is a records question a coastal State can answer from its own files.

Why the radio layer permits this

AIS was never built to resist this. ITU-R M.1371 provides the technical characteristics of an automatic identification system using time division multiple access in the very high frequency maritime mobile band – a cooperative broadcast, with no authentication of the transmitter and no integrity protection on the payload ^[7]. The IMO’s carriage requirement frames the purpose plainly: transponders provide position, identification and other information about the ship to other ships and to coastal authorities automatically, for safety of navigation ^[8]. A system designed so that vessels can see each other is not a system designed so that a State can prove where a vessel was.

The supporting instruments reflect that. The US Coast Guard’s reference set names ITU-R M.1371 as the primary AIS standard alongside IMO performance requirements in MSC.74(69), operational guidelines in A.1106(29), and MSC.347(91), which recommends that administrations ensure the integrity of the radio channels used for AIS in their waters ^[6]. Channel integrity is spectrum management. It is not message authentication, and nothing in that set supplies message authentication.

Spoofing is also not the only way to break an AIS identity, and this is the point most often missed. Windward’s Q2 2026 submarine cable risk report describes a sanctioned tanker loitering above Mediterranean cables that was engineered to be invisible to AIS-only monitoring through reflagging and reissuance of its maritime mobile service identity ^[10]. No false transmission is required: the identifier is reassigned administratively, and a monitoring system keyed to it loses the vessel while it continues to broadcast honestly. The same report records 34,081 slow-speed vessel activities near global submarine cable infrastructure in Q2 2026, with activity attributed to states of concern falling nearly 58 per cent quarter on quarter, from 8,042 events to 3,393 ^[10]. Those count behaviour near cables, not interference with them.

What to do

• Treat a maritime mobile service identity as a claim, not an identity. It is asserted by the transmitter, carries no signature, and can be reassigned administratively ^[7][10].
• Preserve raw, per-receiver AIS records with source attribution and timestamps, not merged tracks. Where an impossible track is the evidence, only the unmerged records distinguish manipulation from a transmitter fault or an aggregation error ^[5].
• Corroborate before concluding. A position claim becomes an observation when an independent sensor – radar, patrol aircraft, satellite imagery, or acoustic detection – agrees with it. The PCG’s own aircraft challenge is that corroboration, and it is the part of this account that does not depend on the transponder ^[1][4].
• Separate the questions when reporting an incident: was the broadcast false, was the activity research, and was consent given. The first is a signals question, the second a behavioural inference, the third a documentary one under UNCLOS Articles 246 and 248 ^[5][9].
• For operators of cable landing infrastructure, treat AIS-only monitoring of approaches as incomplete by design, and budget for a second sensing modality rather than a better AIS feed ^[10].

The R3KONX view

Electronic intelligence is partly the discipline of knowing what a signal can carry as proof. An unauthenticated broadcast can prove inconsistency – two positions that cannot both be true – and it cannot prove identity, intent or location. The Batanes case is a clean demonstration of both halves at once, and the PCG’s handling of it, sending an aircraft to look rather than resting on the feed, is the right instinct.

For this region the exposure is structural rather than episodic. The waters off Batanes sit on cable routes carrying a large share of the traffic between North and South East Asia, and the monitoring picture for those approaches rests on a cooperative broadcast specified in 2001 ^[7]. Malaysia’s own cable approaches, in the Strait of Malacca and off Sarawak and Sabah, depend on the same instrument. The question to take from this incident is not whether one vessel lied. It is what a maritime authority could prove if it had to, and whether the records it keeps today would survive the scrutiny this case is now getting.