Eleven hundred ships in one day
On February 28, as the war with Iran began, electronic interference disrupted navigation for more than 1,100 commercial ships in Emirati, Qatari, Omani, and Iranian waters. That is Windward's count. Lloyd's List Intelligence logged 1,735 separate interference events against 655 vessels in the first four days, with daily incidents nearly doubling from 350 to 672.
The strangest part was where the ships appeared to be. Navigation systems placed tankers at airports, at a nuclear power plant, and on dry land inside Iran. Some tracks drew slow circles in open water. The ships were where they had always been. The reports of where they were had been taken over.
This is now the normal environment
The Gulf was not an outlier. GPS signal loss events affecting aircraft rose 220 percent between 2021 and 2024 according to IATA data. In June, Ukraine lost contact with four uncrewed surface vessels roughly 250 kilometers off the Romanian coast. One of them turned up at a berth in the port of Constanța the next day and detonated after the area was evacuated. Ukraine assessed Russian electronic warfare as the most likely cause.
Any unit planning to operate in the Baltic, the Black Sea, the Middle East, or the Western Pacific should assume the same thing. Satellite positioning will be degraded, and some of the positions it reports will be wrong on purpose.
A position is a claim
Most tactical software treats a reported position as a fact. A track arrives with coordinates, and the map draws it there. That design made sense when spoofing was rare. It is now a vulnerability, because whoever controls the report controls the picture, and the operator looking at the map has no way to tell.
The fix is to treat every position as a claim made by a source. A claim has an origin, a time, and a level of trust. Claims from independent sources can corroborate each other or contradict each other. A track built on one unauthenticated broadcast should look different on the map from a track confirmed by a camera, a radar return, and a friendly report.
What the software has to do
In EdgeLance every observation keeps its source, its timestamp, and how much it should be trusted, all the way from the sensor to the map and into the record. A ship that jumps forty kilometers inland in one update gets flagged as an anomaly instead of drawn at its new position. When two sources disagree, the operator sees the disagreement instead of a silent average.
None of this needs a connection to anywhere. Corroboration happens on the team's own devices, using the team's own sensors. That matters most in exactly the places where spoofing is worst, because those are also the places where the link to higher is least reliable.
Corroboration gets better with every modality the team brings. A camera that sees a hull, an RF receiver that hears its radar, and a lookout's report are three independent claims, and a sensor fabric that fuses six modalities can weigh them against a transponder that says something else. Cloud-dependent AI has the opposite problem, because contested spectrum takes the cloud away at the same moment the positions start lying.
Trust the picture you can explain
A commander does not need a perfect picture. A commander needs to know which parts of the picture to trust. Software that can say "this track rests on one broadcast nobody has confirmed" is worth more in contested spectrum than software that draws every report with the same confidence. The adversary is already counting on the second kind.
The controls that keep a bad report from becoming a bad decision are on our security page. The short version is that every report keeps its source from the sensor to the record, and the picture never forgets where a track came from.