In partnership with
Dear {{ first_name | reader }},
On the evening of 15 March 2026, an unidentified luminous object appeared in the approach path to Cát Bi International Airport in Vietnam.
Two separate flight crews reported it. Four inbound flights diverted to Nội Bài. Two departures sat on the ground. Operations resumed after several hours of monitoring and verification, at 00:35 the following morning.
The object was never conclusively identified.
This is almost a perfect illustration of the communication problem that follows when you do exactly the right thing operationally. You detected an unknown. You acted. You protected the airspace. And then you had to explain all of that to a public that experienced the delays, saw nothing, and was told nothing conclusive.
That is the crisis communication challenge of the drone era. Not the object. The aftermath of appropriate caution, the necessary response to something you still cannot name.
I have been meaning to write this piece for some time. Cát Bi gave me the clearest example I have seen.
Enjoy!
Philippe
WAG THE DOG NEWSLETTER | ISSUE WEEK 19, 2026
KEY TAKEAWAYS
The identification problem is the communication problem. Most drone incident plans prepare for a confirmed drone. The Cát Bi object was never confirmed. Most plans have nothing to say about that moment – when you acted correctly, the threat remains unclassified, and the public is still waiting for an explanation. That gap is where credibility is won or lost.
Build a shared vocabulary before the incident. "Reported sighting," "suspected UAS," "confirmed incursion," "hostile act" – these are not interchangeable. Your spokesperson cannot improvise that taxonomy at 11pm with two flight crews reporting a luminous object and three journalists on the phone. Write the definitions. Train to them. Get every agency in the coordination chain using the same words.
Issue a first-hour holding statement with four fixed slots. What was seen. What action was taken. What is not yet confirmed. When the next update comes. The structure buys you forty minutes. The complexity is not writing it – it is having it agreed in advance.
Pre-agree spokesperson roles by audience, not by agency. Aviation safety, law enforcement, national security – these are audiences, not departments. Decide before the incident who speaks to whom. The agency that owns the operational response does not automatically own the public narrative.
When the object remains unidentified, explain the system that responded to it. Credibility is not rebuilt by delivering a conclusion you do not have. It is rebuilt by explaining that the procedure which diverted four flights for an unidentified luminous object is the same procedure that would divert them for a confirmed hostile drone – and that is precisely how it should work.
Table of Contents
The incident
On the evening of 15 March 2026, two separate flight crews reported an unidentified luminous object in the approach path to Cát Bi International Airport in Hải Phòng, Vietnam.
One crew placed the object roughly 8 kilometres from the runway threshold at approximately 300 metres altitude. Air traffic control responded. Four inbound flights diverted to Nội Bài. Two outbound flights delayed. Operations resumed at 00:35 the following morning, after several hours of monitoring and verification.
The object was never conclusively identified.
Stay with that for a moment, because it contains almost everything you need to understand about drone crisis communication.
The operational response was not wrong. It was correct, proportionate, and legally required given what two independent flight crews reported: an unidentified object crossing the approach corridor of an active international airport. You act. You divert. You verify. You get an answer if the evidence supports one. In this case, it did not.
And now you have a different problem entirely.
The identification problem is the communication problem.
The planning gap
Most drone incident plans do not grapple with that. They prepare for a confirmed drone. They write holding statements about "suspected UAS activity." They map spokesperson roles and coordination chains.
They do not prepare for the operational requirement to act before identification is possible – and then explain that sequence to a public that experienced hours of disruption and received no conclusive answer about what caused it.
Cát Bi is not an edge case. It is a category.