Three hundred feet. That's the margin. The 737 came out of that dive with roughly three hundred feet between it and the ground, and the only reason anyone on board is alive is that a wounded captain opened a door from the wrong side.
And that near miss, on the thirtieth of September, is now sitting in the middle of an Israeli election campaign, a debate about the entire architecture of aviation security, and a very specific question about whether the pilot's nationality should have kept him off that airplane in the first place.
Which is what Daniel wrote in about this week. Flydubai 1073, Dubai to Tel Aviv, first officer attacks the captain with the crash axe. Daniel wants to know, roughly in this order. One, in the ordinary course of affairs, are flight crews actually subject to vetting by the countries they fly to? Two, is there any precedent for route-specific crew vetting, anywhere, ever. Three, if Israel's response is to build a high-security regime where crews have to be vetted before they're allowed on routes serving Israel, is there anything to suggest that arrangement would actually work?
And four.
He also argues, and I want to give him credit because he argues it well, that the whole idea strikes him as operationally strange. How does that even work with rostering? A foreign carrier suddenly has to submit its employees to a background check regime, and only the ones who pass get the route. Would any carrier tolerate that? And if one did, what does it say to its own pilots. We trust you to fly our airplanes, just not into this one country.
That's the sharpest version of the objection I've seen put, and it's a management question more than a security question.
And the fifth piece. Put the extremist posts to one side, he says. Why are people in Israel gravitating so hard toward the fact that the co-pilot was Omani, and that Oman and Israel have no official relations? What are the problematic implications of that state of affairs.
Five questions, one incident, and an election in three weeks.
So let's do what Daniel asked. Start with the baseline. Who actually vets a flight crew?
Nobody, in the sense you're asking. In the ordinary course of affairs, a flight crew flying into a foreign country is not vetted by that country. The vetting happens at the other end. The pilot's own state and the pilot's own employer carry that burden.
Which sounds like a dodge until you realize it's the entire system.
It is the system. It's been the system since Chicago in 1944. The state that licenses the pilot certifies the pilot, the airline that employs him checks him, and everyone else relies on those two things being done properly. Medical certification to ICAO standards, with mental health screening varying by jurisdiction. Background and criminal history checks tied to an aviation security identification card. Every country has its version.
And what does that background check actually catch?
Mostly a criminal record. That's what it's designed for. And if there isn't one, there may be nothing to flag. This is the point the two CQUniversity researchers made, Leib and Carim Junior, writing in The Conversation on the first of October. Background checks are often limited to criminal history, and there's no broad regulatory requirement anywhere to treat the pilot as a potential threat. Crew training focuses on collaboration in the cockpit, not on the pilot as the threat.
So the mental model the whole industry is built on is that the person in the left seat is a colleague, not a suspect.
Correct. The cockpit is a small, high-trust, high-stakes environment. You can't run it if the two people in it are treating each other as hazards.
So Daniel's first question, answered. Destination countries don't vet. The flag state and the employer do. Now the interesting part: on the fourth of October, Israel effectively stated that out loud.
The government confirmed what everyone in the industry already knew. Israel does not independently vet foreign pilots. It relies on the international airline confirming that it has vetted its pilots to international standards.
And that admission is itself the story, because the criticism is that Israel was supposed to be the exception. Israel is the country everyone assumes does the paranoid thing, and it turns out it was doing the normal thing the entire time.
Netanyahu ordered a thorough review of security procedures for international flights into Israel that same day. And there was already a live argument between the Ministry of Transportation and the Shin Bet about who was responsible, with the ministry saying it doesn't have the resources to conduct security vetting on pilots assigned to fly to Israel. That argument predates this attack.
Which is important. This didn't create the dispute. It exposed it.
The gap existed and everyone in the room knew it existed.
Now the second question. Precedent. Has any destination country ever vetted foreign airline crews as a condition of flying a route to it.
Not in the way he's describing. There is no established precedent for a destination country individually vetting foreign airline crews as a precondition for a route. What exists instead is mutual reliance. The flag state vets, the employer vets, and the receiving country accepts that certification. There are bilateral safety arrangements, data-sharing arrangements, no-fly lists that apply to passengers. Crew is different.
There's no mutual recognition treaty for crew vetting the way there is for airworthiness.
No. We have spent decades building mutual recognition of certificates of airworthiness, of licences, of maintenance standards. All of that is heavily regulated and heavily audited. The vetting of the person is left to the home state.
So the closest thing to a precedent is Israel's own arrangement. Not with a foreign country broadly, but with Flydubai specifically.
Which is the piece I think most people have not fully absorbed. Under its agreement with Israel, Flydubai was already obligated to exclude Omani pilots from Israeli operations, because Oman and Israel have no diplomatic relations. That obligation predates the incident. And the man who attacked the captain was an Omani national flying the Israel route. So the safeguard was not missing. The safeguard was bypassed.
That's the whole thing. The rule existed, and the person the rule was written for flew the route anyway.
And it reframes the question Daniel is asking. He asks whether there's precedent for a route-specific vetting regime. The more pointed answer is that Israel already had a route-specific exclusion, narrower than vetting, easier to administer, nationality-based, and it did not hold.
Which tells us the failure was not that nobody wrote the rule.
The failure was enforcement. And intelligence sharing, which is the deeper problem. The Wall Street Journal reported on the second of October that Oman had previously banned al-Hammami from flying because of radical ideological views. What is not clear is whether Oman shared that with any other state, or with any airline, or with anyone at all. If a state knows one of its licensed pilots is a problem and doesn't tell anyone, then any vetting regime built downstream of that state is building on sand.
And separately, ABC reported that extremist material was found in his possession while he was at Oman Air in twenty twenty-four.
And he stayed in an administrative role. So the state had multiple data points, and the data points did not travel. That's the mechanism. Not the absence of rules. The failure of information to move from the place that knew to the place that needed to know.
Now the third question. If Israel institutes a high-security regime by which flight crews are vetted before being approved on routes, is there any precedent to say that arrangement might work.
There isn't, and the people who study this professionally are not optimistic. Leib and Carim Junior caution specifically against reactive, poorly thought through protocols, because in their words these may cause more harm than help.
That's a lovely sentence. Cause more harm than help.
Unpack it. Screening is expensive. It could deprive airlines of crew, because crew availability is already the binding constraint on a lot of routes. Pilots may be reluctant to fly routes with heavier screening, which creates a soft strike nobody authorizes. And the one that matters most: pilots may conceal physical or psychological problems because they fear being grounded. That last one is the failure mode that actually kills people. If you make the consequences of disclosure severe enough, you stop getting disclosure.
That's an argument doctors have been making about pilot mental health for a decade.
Since Germanwings, at minimum.
And it connects to Daniel's rostering point. He says a route-specific regime means the carrier is telling its pilots: we trust you to fly our aircraft, but not into this country. What does that do to a cockpit.
It creates a two-tier trust structure inside a single airline. The first officer who is completely trusted on the Dubai to Riyadh rotation is suddenly a suspect on the Dubai to Tel Aviv rotation. That's not a secret you can keep in a crew room. It surfaces. And either way the cockpit stops being a high-trust pair.
The thing that's interesting to me is that a carrier might actually agree to this.
Some might. If the route is commercially valuable, a carrier will grumble and sign.
But then the pilots are the problem, because the pilots are the ones who have to actually do it. And unions are the ones who negotiate the terms.
Safety culture literature is very clear on this. You cannot impose a trust regime by contract on the front line. The front line has to internalize it or it doesn't exist. So you can get the paperwork signed, and you can have a compliance manual that describes a beautiful system, and the system can be hollow.
Now there's a threshold question, which Daniel doesn't ask but I want to raise, because it sits underneath all of it. Assume everything works. Assume Oman hands over the data, Flydubai runs the check, Israel reviews it, and the crew is clean. How many of the people who would do this are people who have no record anywhere.
That is the right question, and the answer is not good. The overwhelming majority. Someone on a clean trajectory, no criminal history, no extremist posts, no administrative action, who radicalizes privately. That person has no data trail. Germanwings is the clearest example. Andreas Lubitz had sought treatment and kept it quiet. The whole reason the Germanwings story is the reference point is precisely that he was inside the system, passing every check that existed, and the system was designed to keep him safe from passengers.
So Daniel's instinct that this is operationally strange is correct, and it's correct for a deeper reason than he gives.
Not only is it strange to administer. It also gives you a thin veneer of protection that you could mistake for real protection. A regime that screens two percent of the risk and presents itself as screening all of it is worse than no regime, because it changes behavior. You stop doing the things that catch the things the screen can't catch. The colleague who notices something and mentions it. The scheduling supervisor who thinks, that swap looks wrong. Those are the catches that have worked.
And those aren't rules. They're culture. And culture is not what you get out of a compliance manual.
You get it out of years of a workplace where speaking up is safe.
Let's move to the Omani question, because it's the one Daniel circles three times, and he's right to.
The 'why are people gravitating to it' question.
Yes. On its face, al-Hammami's nationality is one biographical detail among many. Why is that fact carrying the political weight of the entire scandal.
Because it converts a story about a single bad actor into a story about a system. If the pilot is a lone wolf with no institutional footprint, the story is a tragedy. If the pilot is a national of a country that individually has no diplomatic relations with Israel, and there was a rule engineered precisely to keep pilots of that nationality off the route, and the rule was not applied to him, then the story becomes about compliance, oversight, and intelligence sharing. It becomes a failure story. And failure stories are the ones that move elections.
The nationality is a hinge.
It's the thing that connects the individual to the institution. Without it, there's no institutional failure to investigate. With it, the Shin Bet, the Ministry of Transportation, Flydubai, and the government of Oman are all implicated.
And there's an uncomfortable asymmetry in there. The state that gets blamed is not just Oman. It's also Israel, for relying on self-certification, and it's Flydubai, for breaking its agreement. Which is exactly why the story is politically hot. It has multiple hosts to pass around.
And the politics is live. Israel's legislative election is on the twenty seventh of October. Netanyahu had warned of an attack plot on the twenty ninth of September; Lapid dismissed the warning as fearmongering. Then Flight 1073 happened the next day. Lapid said nothing like FZ1073 had come up in his security briefing. Opponents are drawing October 7 parallels. The attack is a flashpoint with three and a half weeks to go.
That's why people are gravitating to the Omani fact. Not because it's the biggest fact. Because it's the fact that political actors can do something with.
And the timing is not incidental. An attack that produces headlines three weeks before an election is an attack that becomes an argument.
Which gives the Omani angle a strange kind of double exposure. On the one hand, it points at the real institutional failure. On the other, it gives political actors an easy narrative, and the easy narrative is often the wrong one.
The easy narrative here would be: this is Oman's fault, and we need to punish Oman. Which is plausible-sounding and doesn't help.
Because Oman is not going to be punished by an Israeli election. The relevant actor here for the fix, if there is a fix, is the United Arab Emirates, and the Emiratis have already responded.
They have. Israel suspended all Flydubai flights for six days. The UAE agreed to Israel's security demands. El Al, Arkia, and Israir resumed Dubai flights, which had been restricted for about six months previously. So the practical response has been bilateral and commercial, not a new global vetting regime. The system is already self correcting via commercial pressure and reciprocal arrangements, which is the way it usually works.
There's another piece of the aftermath that I want to raise because it's uglier and it goes to the Omani question. Within sixty seven minutes of the squawk, the false flag material started.
Within sixty seven minutes of the emergency squawk. Over three thousand posts in twelve hours claiming Jewish or Israeli involvement, roughly four thousand using the phrase false flag. So before the aircraft had even been pulled off the runway at Tabuk, the story had already been rewritten by people who were never going to accept the real one.
And that's its own kind of vetting failure. Not of a pilot. Of an information environment.
It's the thing that makes the political response harder, because the political response is happening inside a noise floor that has been lowered artificially.
Let me raise the Germanwings comparison, because it's the one that fits.
It's the closest precedent for the threat model. In March twenty fifteen, Andreas Lubitz, a co-pilot, deliberately flew a Germanwings Airbus into a mountainside in the French Alps. One hundred and fifty people died. He had been treated for depression and had been declared unfit for work by a doctor for a period before the crash, and he concealed the note. The industry's response was a two-person cockpit rule on many carriers, and a conversation about mental health screening that has not resolved.
And it hasn't resolved because the Germanwings failure and the Flydubai failure are the inverse of each other.
They are exact inverses. Germanwings was a pilot who passed every check and was trusted by everyone, and the failure was inside the cockpit. Flydubai 1073 was a pilot who was flagged in advance by one state, and the flag didn't travel, and the failure was between institutions. One is a failure of trust inside the room. The other is a failure of trust between rooms. And the industry does not have a framework for either.
Post 9/11 we hardened the cockpit door.
Bulletproof, reinforced, locked from inside, not openable from the cabin. That's a defense against the threat outside the cockpit. The threat inside it has no equivalent architecture.
Something's been bugging me about the whole debate. Both sides are treating this as a rule problem. Daniel's side says the rule would be unworkable. The government's side says we need a new rule. Neither of them is asking whether the rule that exists is being enforced, and the answer to that is visibly no, because the one rule that would have prevented this exact outcome was already in place.
That's the take. Enforcement and intelligence sharing. The Conversation authors land on it too.
I'm saying it because I think it gets buried in the coverage. Everyone wants to invent a new control, and almost nobody wants to audit the one that already failed.
Because inventing a control is easier than auditing one. Auditing is unglamorous. Auditing says: somebody had to check a list, and the list had a human in the loop, and the human had a bad night. That's not a story you can campaign on.
So we should say what the mechanics actually are. How were Omani pilots kept off Flydubai's Israel routes before this.
On the information available, it was a listed exclusion inside the carrier's own scheduling. The Israel agreement required it, and the way it was implemented was through the rostering system. Under that arrangement, if a crew member is obligated to be excluded, and the scheduling software has to know that, then the safeguard is only as reliable as the flag in the system.
And if the flag is a remark rather than a hard constraint.
Then a scheduler has to notice it. And if the scheduler is building a rotation to a route that operates daily, with sick calls and swaps and last minute changes, then a remark is not a piece of software. It's a hope.
It also fits the fact that he had recently transferred from Oman Air to Flydubai, which means his record was moving between employers at the same time.
And the employer that hired him did not have visibility into the flag the previous employer and the previous state had. That's the direction of information failure. It runs against the flow. The thing that knew was upstream, the thing that needed to know was downstream, and the pipe was not plumbed.
Which is arguably the biggest single finding of the week. It's much less dramatic than a new vetting regime, and it's much more useful.
There's a roster shape to it. Flydubai runs Dubai to Tel Aviv as a daily. That means roughly one crew sequence a day, which is a few hundred a month. Within a fleet of hundreds of pilots, you need the exclusion to hold on every single one of those assignments. Every.
Ninety nine point nine percent is a failure at that volume.
Over a year, a hundred percent minus one is roughly three roster events where someone has to notice.
You only need one.
You only need one.
Where does that leave the Israeli proposal, then. The government's review is live, the pressure is on, and the election is in three weeks.
If the review lands on 'build a new vetting regime', they will run into the same class of problem they already have. Anything they build will be a system that depends on humans reading flags and making decisions, in a scheduling context that is already fast, distributed, and full of exceptions. There is no version of this that is fully automated, because the underlying data lives with foreign states and foreign carriers, and they can only share what they choose to share. You can automate the pipe. You can't automate the trust.
You can make the pipe compulsory.
In theory. But then you're back to Daniel's objection, which is about what a carrier is willing to sign.
Or you route around foreign carriers entirely.
You do. You fly El Al. And El Al is not just a commercial choice, it's a security choice. Israeli crews, Israeli vetting, Israeli threat model. That is the solution Israel actually has available to it, and it's been the solution for decades. It doesn't scale to every route, and it doesn't fix the problem for the crews Israel cannot control, but it does remove the class of threat that this incident represents for the routes El Al flies.
The real question for the Israeli review is not 'can we vet every foreign pilot', it's 'how much of the route network are we willing to shift to carriers we actually control, and at what cost'.
That is a much less satisfying question, which is why it's a much better question.
Okay. I want to get into what happens when these reviews actually run, because it's the part of this debate that almost never surfaces. The rule exists, the rule failed, and now there's a proposal to write another rule.
The pattern is familiar. The new rule gets written, the compliance document looks solid, the audit trail is impressive on paper, and then the roster is built at two in the morning by somebody with a sick call to cover in the next four hours. The doc says what to do. The pressure says do something else.
I keep coming back to something that's easy to underestimate. A foreign captain from a non-allied state is trusted with a plane full of Israelis a hundred times a year. That trust is not written down anywhere. It's not...
Hilbert: Yeah, that's basically right. The doc isn't the control. The control was one line in a scheduling file and a duty officer who had to notice it.
Hilbert.
Hilbert: This was a while ago. I was doing crew scheduling for a regional carrier out of a base in the Gulf. Small operator, a couple of aircraft doing nothing but shuttle work.
You had nationality restrictions.
Hilbert: We had TLV on the board. And a note that said do not assign to TLV for two passport types. It was not in the system. It was in a spreadsheet that we kept open on the second screen. Every roster push, you had to look at it. And every roster push, the honest answer is sometimes you didn't.
And nobody flagged it.
Hilbert: Nobody flagged it because nobody was told it was a hard stop. That's the thing the write ups miss. A rule can be written, signed off, filed, and then sit in a spreadsheet and be quietly ignored by people who never knew it mattered. The scheduling system does not understand nationality the way you'd want. It understands base, and type rating, and duty limits. Everything else is human. And when the roster's tight, the human reads past it. I watched that happen. Not with an Omani, and not on TLV. But the mechanism is the same.
The fix is not to write a better rule.
Hilbert: No. The fix is to make it impossible to ignore. Hard block in the software, not a note. Cross check the passport database against the assignment, not against the scheduler's eyeballs. If the policy can be defeated by a tired person looking at the wrong screen at two in the morning, then you don't have a policy. You have a preference.
That's a useful distinction. Preference versus control.
Hilbert: There's a second thing. The information about him existed in Oman. The scheduling team at his employer probably never saw it. Their rule was a note about Omanis. It was not an attempt to evaluate the man. Even a perfect version of that rule would not have caught him if he had dual nationality or a different passport somewhere else in the rotation. What would have caught him is Oman telling the UAE, and the UAE telling the airline, and the airline's data system flagging the individual by identifier, not by citizenship. That's the pipe. It's harder than it sounds because the data sits in systems that don't talk to each other, in regimes that don't share, with legal frameworks that were written for a different threat model. It took us two years to get our own HR system to flag visa expiry. Visa expiry. So I'm skeptical of a sixty day review.
That's fair. I don't think either of us thought a new regime solves the problem. I think we thought it might paper over it.
Hilbert's right about the pipe. And there's a concrete data point that supports him. Under the Flydubai agreement, the exclusion was nationality-based. Al-Hammami was an Omani national. So the rule should have caught him. It should have caught him by the cleanest possible test, citizenship. And it didn't. Which means the rule was not bypassed by a clever move by the attacker. It was bypassed by the mechanics of a scheduling system that didn't treat the rule as load-bearing. That's the actual finding.
That is what an Israeli review should be looking at. Not whether to write a new regime. Whether the existing regime, such as it is, can be hardened. It can be. The mechanisms exist. Whether they will be is a different question.
Which I think is the point Daniel is circling. His instinct that Israel vetting foreign crews is untenable is right. And the more precise version of that is: it can't work because the model is mutual reliance, and the mutual reliance is what failed. The fix has to be a fix to the mutual reliance, not a replacement of it. And that is an intelligence and data problem, not a security regime problem.
Leave the listener with something to chew on. I'll take the closing thought. The insider threat is the last hard problem in aviation security because it inverts the model we've spent twenty five years building. We hardened the cockpit door so a threat from the cabin couldn't get in. Now we have to consider what to do about a threat that walks through the door every day and sits down in the right seat.
The honest answer, at the moment, is that the industry does not have a framework for it. The one that exists is a nationality-based exclusion in a scheduling system. That is not a solution. That's a hope.
That's the question to hold. Not what new rule to write, but whether the rules already written are being enforced. Because on the current evidence, at least one of them was not.
Before we go, our producer Hilbert Flumingtop, who has more to say about two in the morning than most.
This has been My Weird Prompts.
If you like what you heard, a review helps more than you'd think. We read every one.
We'll be back soon.
