A Boeing 737 MAX 8 cruising at thirty-four thousand feet over the Jordanian border, and then, in under sixty seconds, it's at seventeen thousand. That's a descent rate of over four thousand meters a minute. And the reason it pulled out of that dive is that a passenger in the front row heard screaming through a cockpit door that had just popped open.
And the man who opened that door was the captain, bleeding, having just been stabbed by the person sitting next to him.
That's the part that stops you cold. Daniel sent us a whole thing about yesterday's Flydubai flight from Dubai to Tel Aviv, and he's working from the early wire copy, which got the story almost exactly backwards. He wrote that the pilot attacked the co-pilot, that the co-pilot managed to open the cockpit in his last moments of consciousness, that civilians commandeered the plane and landed it in Saudi Arabia. He's asking whether this was an insider threat kamikaze attack aimed at Israeli nationals, why the insider threat model is so hard for regulators even after everything we rebuilt post-9/11, how air safety has actually evolved to deal with this, and whether this is the first incident of its kind. He also asks about Israel's controversial security code system for approaching aircraft.
And the correction isn't a small one. The first officer attacked the captain. The captain is the one who got the door open. And the plane was landed by off-duty Flydubai pilots who happened to be aboard, not by the passengers alone.
So before we dive in, we need to fix the record, because the real story is more remarkable than the version Daniel had. Let's start there.
Flight 1073. Boeing 737 MAX 8, registration A6-FKF, in service since December 2022. Departed Dubai International at 7:05 a.m. Gulf time on the thirtieth of September, bound for Ben Gurion. One hundred and eighty people on board. A hundred and seventy-four passengers, six crew. Roughly a hundred and fifty of those passengers were Israeli citizens.
Which is the detail that makes the Israeli government's framing make sense.
Right. Around 9:21 a.m., near the Jordan–Saudi border, the aircraft plunged from about thirty-four thousand feet to seventeen thousand in under a minute. It briefly climbed back to twenty-one thousand seven hundred and fifty, then went down again to fifteen thousand and turned back into Saudi airspace. The transponder squawked seven seven zero zero, general emergency. Then it squawked seven five zero zero, which is the code for unlawful interference. Hijacking.
And that's when the Israeli Air Force scrambled.
They'd lost contact and they had a hijack code. Fighter jets up. Communication was restored, the relief crew declared an emergency at 9:44, and after about an hour at fifteen thousand feet the plane landed safely at Prince Sultan bin Abdulaziz Airport in Tabuk at 10:45 a.m. Gulf time. Zero fatalities. Six or more injuries. Both pilots hospitalized, the captain initially in serious condition with deep cuts to his head and hands.
Now the cockpit itself. Walk me through it, because this is where the wire copy fell apart.
The first officer stabbed the captain. An apparent attempt to hijack or crash the aircraft. The captain, wounded, managed to open the cockpit door. That's the hinge of the whole thing. A passenger was waiting near the lavatory, saw what was happening, told a flight attendant and her husband, Tzvika Manes, a Bank Leumi executive sitting in the front row.
Manes saw the captain fall out of the cockpit with the first officer on top of him, stabbing him, as the plane went into the dive.
Manes went in. With an off-duty pilot and other passengers. Yaniv Hayun, Musaev Shota, Asaf Rajuan. They overpowered the first officer, who was allegedly trying to disable the flight controls, and got him out of the cockpit.
Hayun is a plumber.
Hayun is a plumber, and he pulled back on the yoke to level the plane. He later said he knew to do that from watching the docudrama series Mayday. Air Crash Investigation.
A plumber who watched enough Air Crash Investigation to know which way the yoke goes when the nose is down. That's not a training program. That's a man who paid attention to television.
The off-duty pilot then took over and flew the aircraft. The first officer was restrained with wired headphones, which is the detail I keep thinking about. Shota, the dentist, treated the captain's injuries. And then there's the part that hasn't gotten nearly enough attention.
The rudder.
Photographs after landing show most of the aircraft's rudder was missing. Likely detached during that rapid descent.
Which means the airplane that landed at Tabuk was missing a primary flight control surface, and the people flying it may not have known.
They would have felt it. Asymmetric control authority, degraded yaw response. But whether they understood what they were feeling in the moment, I doubt it. You're an off-duty pilot who just watched a colleague get stabbed, flying an aircraft you didn't brief for, and the rudder is partly gone. That's not a scenario anyone rehearses.
Four thousand meters a minute. What kind of aerodynamic load does that put on a control surface?
It's enormous, and I want to be careful here because I don't know the failure mode. The rudder could have been damaged in the struggle, it could have been overstressed in the dive, it could have departed later. But a rudder is a big, lightly built structure. It's designed for normal aerodynamic loads with margins, not for whatever happened in that cockpit. The investigation will spend months on it.
So the sequence is: co-pilot attacks captain, captain opens door, passengers subdue attacker, plumber levels the aircraft, off-duty pilot flies it, dentist treats the captain, and they land in Saudi Arabia with most of a rudder.
And the first officer was arrested by Saudi authorities on landing. Passengers and cabin crew were questioned for seven hours.
Seven hours. What did the Saudis say?
Their initial assessment was that the first officer intended to deliberately crash the plane. Motive unknown. Unclear whether he acted alone. Israeli officials examined whether he intended suicide. Defense Minister Israel Katz called it an attempted jihadist terror attack. Netanyahu said it was too early to establish who was behind it and raised the question of Iranian involvement.
Raised the question. Didn't answer it.
The investigation is multinational. Israel, the UAE, Saudi Arabia, and the United States. On the first of October Netanyahu said UAE President Mohamed bin Zayed agreed to let Israel participate.
So let's get to what Daniel was actually circling. The insider threat. Why is this so much harder than everything we built after 9/11?
Because everything we built after 9/11 was built to keep threats out of the cockpit. Hardened doors. Access control. Screening passengers before they board. The entire architecture assumes the threat is on the other side of the door.
And this threat was in the left seat.
Or the right seat, in this case. The first officer. The person you're supposed to trust with your life while you're in the lavatory. The security model has no answer for that, because the answer would be to not trust the person you're required to trust.
So how do you screen a pilot?
All commercial pilots undergo medical certification under ICAO standards, which includes mental health screening in some capacity. But the rules differ enormously by jurisdiction. And security background checks, the kind that come with an aviation ID card, may be limited to identifying criminal history. If there's no criminal history, there may be nothing to draw from to determine that someone could be a security threat.
Which is a polite way of saying the system looks for a record, and if there isn't one, it shrugs.
And the resources for those background checks vary enormously country to country. A carrier in one jurisdiction may do deep vetting. A carrier in another may run a criminal record check and call it done.
What about training? Do crews train for a colleague turning on them?
Typically not. Crew training is set by individual airlines based on international guidelines, and it typically does not focus on pilots themselves as a potential threat. There's no broad regulatory requirement to address pilot-as-threat scenarios. And there's a reason for that, which is that pilot training emphasizes collaboration. Crew resource management. The whole doctrine is that two pilots work together and cross-check each other. That's the opposite of what you'd need to train for here.
You'd be training for the thing that destroys the thing you're training for.
That's the trap. Steven Leib and Guido Carim Junior, two aviation professors at CQUniversity in Australia, wrote this up on the first of October, and their line is: what does it look like when the threat is inside the cockpit to begin with, from someone we are supposed to trust?
And their warning is that the industry shouldn't overcorrect.
They explicitly say regulators need to be careful about reactively implementing new protocols, because those may cause more harm than help if poorly thought through. Which is worth sitting with, because the obvious response to this is a new rule, and the new rule might make things worse.
Give me an example of how a rule could make it worse.
Rule of two is the obvious one. After Germanwings, European regulators required at least two authorized people in the cockpit at all times. So if one pilot leaves, a flight attendant comes in. That sounds strictly better. But it also means a flight attendant is now in the cockpit during a pilot's mental health crisis, and there's a documented case where a pilot locked the door and the crew couldn't get back in. You've added a person without adding a defense.
And the Germanwings case is the one everybody reaches for. Walk me through it.
Twenty-fourth of March, 2015. Germanwings Flight 4U 9525. Co-pilot Andreas Lubitz deliberately flew an Airbus A320 into the French Alps. All one hundred and fifty people on board died. And the detail that matters for us is that Lubitz waited until the captain left the cockpit, then locked the door. The captain reportedly tried to break it down with an axe. Couldn't get in.
So the hardened door that saved flights after 9/11 is the same door that trapped the Germanwings passengers.
Same door. Same certification. Same design intent. It's doing exactly what it was built to do. That's the paradox in one sentence.
And Flight 1073 is the counter-case. Because in Germanwings, the other pilot was out of the cockpit. In this one, the other pilot was in it, and survived long enough to open the door.
Which is precisely the mitigation that mattered. The captain being wounded but conscious. The door opening. The relief crew being aboard. Those are the things that made the difference, and none of them were a rule.
Is this the first time?
No. And this is where the precedent chain gets interesting. The closest structural analogue is FedEx Flight 705, 1994. An off-duty FedEx flight engineer, facing dismissal, attacked the crew mid-flight with hammers and a speargun. Attempted murder-suicide, intent to crash a DC-10.
Hammers and a speargun.
The crew was badly injured and they fought him off and landed the aircraft. Three men. One of them flew the plane with a fractured skull. That's the structural analogue because it's an insider, it's in the air, and it's a fight in the cockpit that the crew won.
What else is in the chain?
SilkAir Flight 185 in 1997, EgyptAir Flight 990 in 1999. Both cases where pilot suicide or homicide was suspected or found, and both remain contested. Mozambique Airlines Flight 470 in 2013, where the captain deliberately crashed a LAM Embraer 190 in Namibia and killed all thirty-three on board.
So the pattern isn't new. The pattern is old and the mitigations are thin.
And the mitigations we do have are mostly reactive. Germanwings produced rule of two. That's the one clear regulatory response to a pilot-as-threat event, and it's a decade old.
Now, Daniel asked about something specific. Israel's security code system for approaching aircraft. What do we actually know?
Here I have to be careful, because I couldn't find any source corroborating a named IFR security code system. I looked. What is documented is different and more concrete.
Go on.
The first officer was Omani. Oman and Israel do not have diplomatic relations. Under its agreement with Israel, Flydubai was obligated to exclude Omani pilots from Tel Aviv operations.
So the flight may have been non-compliant with a bilateral security arrangement before it ever took off.
That's the reporting. Israeli officials said foreign airlines may not use pilots from countries without diplomatic ties to Israel on the route. Transport Minister Miri Regev called for Flydubai flights to Israel to be halted. Flydubai suspended all flights to Ben Gurion for six days.
And then the UAE did something.
The UAE agreed to longstanding Israeli security demands that had previously prevented Israeli airliners from flying there. El Al, Arkia, and Israir announced they'd resume Tel Aviv to Dubai flights, after a six-month hiatus.
So the practical outcome of this incident is that the UAE gave Israel something it had been asking for, and Israeli carriers get their Dubai routes back.
That's the geopolitical ledger. An attempted mass-casualty event, and the consequence is a security concession and a route resumption.
Which is the part Daniel's original framing was reaching for with the security code question. He was asking about the mechanism Israel uses to vet approaching aircraft, and the honest answer is that the mechanism is bilateral agreements with crew-nationality restrictions baked in, not some named code system.
And the Omani detail is the one that should worry people, because it suggests the agreement existed and wasn't enforced. Which is a different failure than not having an agreement.
Let's stay with that. Because there's a version of this story where the lesson is that we need more rules, and there's a version where the lesson is that we had a rule and it didn't hold.
And the second version is harder, because it means the problem isn't the rulebook. It's the enforcement, and the enforcement is a person at a desk checking a crew manifest.
Now the forward-looking piece, because this is where the Conversation analysis gets uncomfortable. Single-pilot operations.
Right. The industry is actively working toward reduced-crew and single-pilot operations, and eventually remote and pilotless aircraft. And the argument from Leib and Carim is that comprehensive risk assessment is essential as those develop, because the entire premise of single-pilot ops is that you've removed the second human who can intervene.
Germanwings was the case where the second human was locked out. Flight 1073 is the case where the second human was the threat and still saved the day, because he opened a door.
Both cases point at the same conclusion, which is that the second person in the cockpit is load-bearing, and we keep talking about removing them.
That's a strange thing for an industry to be doing in the same decade as Germanwings.
It's a strange thing to be doing in the same decade as a lot of things. But the economics are real. Pilot shortage, cost per seat, automation reliability. The pressure is genuine.
What about the mental health screening side? Because that's where the Germanwings case pointed and where the Flydubai case points.
Lubitz had a history of depression and had been treated, and the medical system knew something. The gap was that the information didn't reach the operator in a usable form, partly because of privacy rules and partly because the pilot concealed it. And that's the same gap here. Medical certification exists. Mental health screening exists in some capacity. But the rules differ by jurisdiction and the reporting pathways are inconsistent.
You can't screen for something that isn't disclosed.
You can't. And you can't build a system that assumes disclosure, because the people who most need to be caught are the ones least likely to disclose.
Which is the same problem as the criminal record check. The system looks for a signal, and the signal only exists if someone already generated it.
Which is why the honest answer to Daniel's question about whether air safety has evolved to deal with this is: partially, reactively, and unevenly. Post-9/11 hardened the cockpit against outsiders. Germanwings produced rule of two in Europe. And the pilot-as-threat scenario still has no broad regulatory requirement attached to it.
Let me put the uncomfortable version of this. The reason there's no requirement is that the mitigation would be intrusive. Continuous psychological monitoring. Peer reporting systems. Restrictions on who can fly which routes. Every one of those has a cost, and the cost is paid by the ninety-nine point nine percent of pilots who are fine.
The cost of the cost is that you get a system that only responds after something happens.
Which is where we are.
Hilbert: Fourteen.
Sorry?
Hilbert: Fourteen years I drove a van for a medical courier outfit out of a depot in Bay Ridge. And there was a guy on that run, Charlie Prine, who'd been flying charter before he took the courier job, and he got pulled off the schedule one Tuesday and nobody ever said why.
Pulled off how?
Hilbert: Just wasn't on the sheet anymore. Dispatcher told us he'd moved to days. He hadn't moved to days. Another pilot, a guy named Delgado, had gone to the chief pilot and said something. Not a report, not a form. Just went in and said he was worried. And Charlie was off the line inside a week, and he was back six months later, and nobody ever knew what happened in between.
The thing that caught him wasn't a screening system.
Hilbert: There was no screening system. There was a guy who'd flown with him for nine years and noticed something was off, and had the nerve to say it to somebody who could act. That's the whole mechanism. That's the only mechanism I ever saw work.
The removal was quiet.
Hilbert: It was quiet because it had to be. You can't put a man's medical history on a crew room wall. So it happens in an office with the door shut, and there's no record anybody can count, and no way to know how often it's happening or whether it's working.
Which is the part that bothers me. If the system that actually catches these people is informal, then we have no idea how good it is.
Hilbert: You have no idea how good it is, and you have no idea how many times it doesn't happen. Charlie had Delgado. Not everybody has a Delgado.
That's the gap between the policy discussion and the actual thing. We talk about screening, and the screening that works is one colleague deciding to walk into an office.
Hilbert: The man who walks into the office is taking a risk. He's accusing a friend of something he can't prove. Half the guys I worked with would have kept their mouth shut and felt bad about it later.
The system depends on the one person willing not to keep their mouth shut.
Hilbert: That's how it worked everywhere I ever was. The paperwork is for after.
Herman, take us out on this.
The thing I keep landing on is that the heroism here is real and specific, and it's also fragile. Captain Machchhar opened a door while bleeding. Tzvika Manes went into a cockpit where a man was being stabbed. Yaniv Hayun knew which way to pull the yoke because he watched a documentary series. A dentist treated a stab wound at altitude. An off-duty pilot flew a damaged aircraft to a safe landing in a country that wasn't the destination.
Every one of those is a person who happened to be in the right seat at the right time.
Which is the uncomfortable part. The mitigations that saved this flight were the relief crew being aboard and the captain surviving long enough to open a door. Neither of those is a rule. Both of those are luck.
The next one may not have a plumber who watches Mayday.
The next one may not have a captain who can still reach the door.
The motive here is still unknown. Whether the attacker acted alone is unconfirmed. Netanyahu raised Iranian involvement and didn't answer it. The structural damage to the rudder is going to be studied for months. And the regulatory question, whether to mandate pilot-threat training and stricter vetting, or to hold off because a bad rule is worse than no rule, is open.
The precedent chain says this isn't going away. FedEx 705, Germanwings 9525, Flydubai 1073. Three cases across three decades, and the only clear regulatory response is a decade-old European rule that adds a flight attendant to the cockpit.
If you found this one worth your time, rate and review us wherever you get your podcasts. It helps other people find the show.
Thanks as always to our producer, Hilbert Flumingtop.
This has been My Weird Prompts. Email us at show at my weird prompts dot com. We'll be back soon.
