On the morning of September 25th, 1997, two men stood outside an office building in Ammon, Jordan, and waited.
They looked like tourists.
They were dressed casually, blue jeans, t-shirts.
They had arrived in Jordan on Canadian passports, which identified them as ordinary civilians.
They had been in the country for a week, moving quietly, doing nothing that attracted attention.
They carried no firearms, no knives, nothing that would register as a weapon to anyone watching.
What they carried was a small metal device wrapped in white cloth that contained a synthetic poison designed to kill a human being within 48 hours without leaving any trace that a poison had ever been administered.
The plan was this.
A man would walk past them.
One of them would step close, press the device briefly to the soft tissue of the man’s left ear, discharge the aerosol, and walk away.
The man would feel something odd, a pressure, perhaps a sound, and then nothing.
He would go into his office.
He would work through the morning, feel unwell through the afternoon, and by evening he would think he was coming down with something.
He would sleep, he would not wake up.
and no doctor, no forensic pathologist, no intelligence analyst would be able to prove that anything other than a sudden severe illness had taken his life.
The operation was authorized by the prime minister of Israel.
It was directed by the director of MSAD.
It was carried out by operators from the Kedon unit, MSAD’s elite assassination element, [music] whose existence Israel did not officially acknowledge.
It was designed to be the kind of thing that simply happened.
A man here today, gone tomorrow.
No gunshots, no blood, no evidence, no diplomatic incident.
Within 72 hours, it would produce one of the worst diplomatic crises in Israel’s modern history.
His name was Khaled Marshall.
And in September 1997, he was not famous.
That is worth holding for a moment because everything that follows depends on understanding what Marshall was and was not at the time Israel decided to kill him.
He had been born in 1956 in Silwad, a village in the Jordanian administered West Bank.
His father was a farmer and an imam.
When the 1967 war came and Israeli forces occupied the West Bank, the family fled, part of a wave of roughly 300,000 Palestinians displaced in what Palestinians call the Knoxa, the setback.
Mashall grew up in Kuwait,
studied physics at Kuwait University, became a teacher.
He was not a military man.
He had not fought.
He had built his standing in the Palestinian Islamist movement through organizational work, recruiting, fundraising, institution building.
In 1992, he became a founding member of Hamas’s political bureau.
In April 1996, at 40 years old, he was elected its chairman.
He ran the political bureau from Ammon, Jordan.
This was Hamas’s external leadership, the apparatus that sat outside the reach of Israeli military and security forces, managing international relationships, coordinating financing, providing strategic direction to the movement as a whole.
It was not a ceremonial position.
The external leadership made decisions that had consequences inside Gaza and the West Bank.
But Marshall himself was not a field commander.
He did not plan specific attacks.
He was in the language of intelligence analysis, a political and organizational node, the man who kept the structure functional and connected to the broader world.
Israel classified him as a primary target.
Anyway, [music] and the reasoning was not complicated.
Hamas’s operational capacity inside the occupied territories depended on the infrastructure the external leadership maintained.
cut the external leadership, particularly its chairman, and you created disruption, uncertainty, a succession struggle at the precise moment the organization was at the height of its operational tempo inside Israel.
That operational tempo, by the summer of 1997, had become intolerable.
On July 30th, at 1:15 in the afternoon, two Palestinian suicide bombers detonated in the Mahane Yehuda market in Jerusalem.
The city’s central open air food market crowded at midday with shoppers, teenagers, tourists.
The bombers positioned themselves 45 m apart in a central alley and detonated almost simultaneously.
16 people were killed.
178 were injured.
The dead included an Arab-Israeli citizen.
Many of the injured were young.
5 weeks later on September 4th, a triple bombing hit the Ben Yehuda Pedestrian Mall in central Jerusalem.
Five more dead, over 190 wounded.
Netanyahu’s government was under severe public pressure.
The security cabinet met.
The question was not whether to respond, but how.
MSAD director Danny Yatam presented options.
Netanyahu made a decision.
The Hamas external leadership would be targeted.
According to subsequent reporting, Netanyahu personally selected Mashaw from a list of candidates.
The authorization moved through the security cabinet and landed on Yatum’s desk as an operational directive.
What happened [music] next inside Mossad’s planning apparatus is where the first problem was embedded, though it would not be visible as a problem until it was too late.
The operation was designed around a single outcome.
Success.
The compound worked.
Marshall died, the agents exfiltrated, the cause of death was obscured, and Israel had decapitated Hamas’s external political leadership without leaving fingerprints.
Every element of the planning served that scenario.
The forged Canadian passport served that scenario.
Canada attracted no suspicion in Jordan, and Canadian nationals had no obvious reason to be Israeli intelligence officers.
The aerosol delivery served that scenario.
No physical wound, no obvious mechanism, nothing for a pathologist to find.
The 48-hour death curve served that scenario.
By the time Marshall died, the agents would be out of the country and the trail cold.
There was embedded in the operation zone materials.
One object that did not fit this framework, the antidote.
MSAD carried the antidote as a safeguard for its own personnel.
If an agent was accidentally exposed during handling, if the device malfunctioned, if there was a transfer of a compound at the wrong moment, the antidote would reverse the effects.
This was not unusual.
Operational precaution in any chemical delivery scenario required that the team have access to a counter measure.
It was responsible planning, but the antidote’s existence meant something that the planning framework did not appear to fully reckon with.
If the operation was ever exposed before Marshall died, the antidote became the single most important piece of evidence in the world.
It was proof, not circumstantial proof, not signals intelligence that could be denied.
Physical proof that Israel had created a specific compound to kill a specific man because only Israel had the specific compounds antidote.
The six-man team arrived in Aman approximately one week before the operation.
They entered through Queen Aaliyah International Airport on different flights from Amsterdam, Toronto, and Paris.
They moved through the city separately, establishing surveillance on Mashel’s movements, confirming his routines, identifying the approach window.
The two active agents, traveling under the names Barry Beads and Shaun Kendall, would execute the hit.
The four support team members would handle logistics, communications, and extraction.
The operation was scheduled.
The target’s routine was known.
The weapon was ready.
The antidote was in Israel.
What the team did not know, what no intelligence assessment appears to have flagged as a serious risk, was that the Jordan operation was being conducted inside a particular geopolitical container that made a failed execution categorically different from a failed execution anywhere else.
3 years earlier in October 1994, Israel and Jordan had signed the Wadi Arab Peace Treaty.
King Hussein had taken enormous political risk to reach that agreement.
Jordan became only the second Arab state to normalize relations with Israel.
The relationship was fragile, new, and personally important to Hussein, who had invested his own credibility in it.
Conducting a covert assassination on Jordanian soil was already a significant provocation.
Conducting one that went wrong that left Israeli agents in Jordanian custody.
That left a dying man in a Jordanian hospital whose doctors did not know what they were treating.
That was a different category of problem entirely.
The container mattered and the operation was not built to survive the container breaking.
On the morning of September 25th, the two agents took up their positions outside Marshall’s office.
His car arrived.
He stepped out, accompanied by three of his sons and two bodyguards.
The bodyguards had already noticed the two men.
What happens in the next 60 seconds will not go according to plan.
And nothing, not the agents, not the prime minister, not the MSAD director in Tel Aviv, will be able to stop what comes after.
[music] To understand what Israel was willing to risk in Aman, you need to understand what the Hamas external leadership actually represented, not as a political abstraction, but as a functioning operational structure that had survived everything Israel had thrown at it inside the occupied territories.
By 1997, Hamas had been a formal organization for a decade.
Shik Ahmed Yasan had co-founded it in 1987 at the outbreak of the first inifat as an armed Islamist movement rooted in the Palestinian Muslim Brotherhood.
From the beginning, it had two distinct functions running in parallel.
a social services network, hospitals, schools, welfare institutions that built genuine popular support among ordinary Palestinians, and a military apparatus that planned and executed attacks against Israeli targets.
The two functions were not separable in practice.
The social infrastructure was the recruitment pool in the political base.
The military wing was the instrument.
Israel had spent years targeting the military wing.
It had arrested commanders, disrupted cells, killed operatives, and Hamas had kept functioning because the military wing was not the organization’s center of gravity.
Its center of gravity was the political and organizational infrastructure.
And the most important part of that infrastructure sat outside Israel’s reach in the offices of the external leadership operating from Ammon, Damascus, and the broader Arab diaspora.
This is the pressure Marshall’s position carried.
He was not a figurehead.
He was the chairman of the mechanism that kept Hamus coherent as a political entity between operational cycles.
He managed relationships with state donors, Iran, [music] Gulf sources, sympathetic governments across the Arab world.
He coordinated the political bureau’s positions on negotiations, ceasefires, and the organization’s public posture toward the Palestinian Authority.
When Hamas’s internal Gaza leadership needed resources, strategic cover, or international legitimacy, it moved through the external bureau through him.
And he did this from Ammon, from the capital of a country that had a peace treaty with Israel, a country whose stability depended partly on maintaining that treaty, a country where Palestinian refugees constituted a significant portion of the population, and where Hamas had a real social and political presence.
That last point is not incidental.
Jordan’s relationship with Hamas was not simply one of tolerance.
Hamas had organizational roots inside [music] Jordan.
It had support among Jordanians of Palestinian origin.
King Hussein had his own complex calculation to manage.
He needed the peace with Israel for strategic and economic reasons, but he could not be seen as simply an instrument of Israeli policy toward Palestinians.
or he risked his domestic legitimacy.
The presence of the Hamas external leadership in Ammon was an awkward reality Hussein had accepted as part of that balance.
What Israel was doing when it authorized the Marshall operation was making a unilateral decision about that balance, deciding that the inconvenience of the Hamas external leadership outweighed the diplomatic cost of acting inside Jordan without Jordanian consent or knowledge.
The calculation assumed the operation would be invisible.
Invisible operations do not create diplomatic costs because they leave no evidence of a decision.
Danny Yetum had come to the Mossad directorship in June 1996 from a distinguished military career.
He was a former commander of the elite Sarat Matgal unit, the same unit Netanyahu himself had served in.
He was trusted by Netanyahu.
He was not an outsider making institutional recommendations.
He was a director whose relationship with the prime minister gave him access and authority and whose background oriented him toward direct action over prolonged collection.
The Mahane Yehuda bombing in July 1997 put that orientation under acute pressure.
MSAD’s role in the postbombing response was not to gather intelligence for a long-term structural campaign against Hamas.
It was to produce results.
The political environment inside Israel demanded visible consequences and the security cabinet wanted to demonstrate that the external leadership was not untouchable.
The bombing had killed 16 people in a market in Jerusalem.
The families of the dead were Israeli citizens demanding [music] that their government act.
Netanyahu was navigating a domestic political environment in which restraint read his weakness.
The targeting of Mashall was in that sense both an intelligence decision and a political one.
The intelligence rationale was legitimate.
Removing the political bureau’s chairman would create a leadership vacuum at a critical moment.
The political rationale was that something needed to happen, something that would not be publicly attributed to Israel that would not produce a visible military response that would give the security cabinet the ability to say in closed session that they had acted.
The operation’s invisibility was not just an operational preference.
It was loadbearing.
It carried the political rationale as much as the intelligence rationale.
If it could not be invisible, it did not serve either purpose.
This is the condition under which the Kedon team arrived in Ammon.
They spent a week in the city before the operation.
7 days of establishing patterns, confirming Marshall’s routines, identifying the window.
Marshall moved with bodyguards, but not with the security apparatus of a head of state.
He used a car.
He walked from the car to his office entrance.
He was accessible in a way that operational planning could work with the approach window.
The gap between the car and the office door was real, repeatable, and brief enough to make Xfiltration plausible.
The support team coordinated logistics across that week.
The four non-executing agents managed communications, the vehicle, the surveillance schedule.
The two active agents, Beads and Kendall, as their Canadian passports named them, refined the timing.
They knew exactly how long Mashel spent between the car and the door.
They knew where his bodyguards typically positioned themselves.
They had selected the approach angle, the moment of contact, the extraction route.
The weapon was technically precise.
The fentinel derivative identified in subsequent reporting as levopentinyl had been engineered specifically for this delivery method.
It was not a field adaptation.
It was a product of Mossad’s technical division designed for absorption through the mucous membranes with a delayed systemic effect that would allow the subject to function normally for several hours before deterioration began.
The aerosol device was small enough to be entirely concealed.
The discharge was quiet enough to require a cover sound, hence the shaken soda can, but under ordinary street noise, the device itself would have been inaudible to bystanders.
The antidote traveled with the team.
This is documented.
It was in Israel at the time of the operation.
The record is consistent on this, [music] meaning that while the operational team had awareness of the antidote’s existence, [music] its physical custody remained in Tel Aviv.
The logic was that the agents themselves were the primary exposure risk during handling and transport.
If an agent required the antidote in the field, it would need to be dispatched.
In a successful operation, this contingency would never arise [music] in a successful operation.
What neither Yatum nor anyone else in the planning chain appears to have modeled with sufficient rigor was the question of physical evidence in a partial failure scenario.
A fully failed operation.
Agents caught before the approach.
No compound administered leaves Israel embarrassed but deniable.
The agents carry forged passports.
They are not carrying weapons by any conventional definition.
The compound is in the device.
Israel can dispute, delay, obfiscate.
A fully successful operation leaves nothing.
Marshall is dead.
The agents are gone.
There is nothing to trace.
But a partial failure compound administered agents caught target alive and deteriorating in a hospital creates a situation with no clean exit.
The agents are in custody.
Their equipment is in Jordanian hands.
The target symptoms will eventually be identified as poisoning [music] and the only question is what poison and the only entity that holds the specific antidote to the specific compound is the entity that designed it.
The partial failure makes the antidote a confession.
No document in the public record, no account by Yatam, no account by Halevi, no finding of the Sichanov Commission that has been made available indicates that this scenario was war gamed at the planning stage with the same rigor as the success scenario.
On the morning of September 25th, Marshall’s car pulled up to his office.
He stepped out.
His sons were with him, two bodyguards.
The bodyguards had noticed the two men waiting.
What happened in the next 90 seconds would answer a question no one in Tel Aviv had adequately asked.
What does this operation look like if it goes wrong in exactly the wrong way in exactly the wrong country with the only proof of guilt sitting in a laboratory in Israel? The bodyguards had been watching the two men for longer than the two men knew.
This is the detail that collapses the operation’s entire premise.
Not a mechanical failure, not a forensic slip, not a communication breakdown between Tel Aviv and the field.
The premise of the Amina operation was that it would be invisible at the moment of execution.
Two men in a crowd, a brief contact, gone.
The premise required that no one with any reason to act on suspicion was paying close attention to the approach.
The bodyguards were paying close attention.
What specifically triggered their alertness is not part of the documented record.
What is documented from Marshall’s own subsequent accounts and from reporting on the operation is that the two men had registered as suspicious before the approach was made.
Whether it was their positioning, their stillness in a busy street, some behavioral tell that trained protective detail personnel are calibrated to notice, cannot be said with certainty.
But the suspicion existed before the device made contact.
The agents moved anyway.
This is the first compounding decision.
The approach was committed.
The window was open.
Marshall had stepped out of the car and was moving toward his office entrance.
To abort at that moment with the target in the window, the bodyguards already watching would have meant withdrawing without executing, burning the week of surveillance, and returning to Tel Aviv to report that the operation had been stood down because the bodyguards looked
suspicious of them.
That report would have gone to Yatam.
Yatam would have taken it to Netanyahu.
A standown based on a protective details general alertness in the absence of any confirmed compromise would have been extremely difficult to justify inside an institution whose culture was oriented toward forward action.
So the approach continued.
The agent closed the distance from behind.
An arm went briefly around Marshall’s neck.
The physical mechanics of pressing the device to the correct location.
The aerosol discharged against his left ear.
At the same moment, Marshall’s daughter called to him unexpectedly from the car.
He turned his head.
The second agent, the one who was supposed to open the shaken soda can to mask the discharge sound, did not open it.
The distraction of Mashel turning had disrupted the timing of the cover action.
The aerosol had discharged, but the contact was not the clean sustained application the operation had been designed around.
Whether a full dose was delivered or a partial dose or something in between, the record does not resolve this question.
What is documented is that enough of the compound entered Marshall’s system to begin a lethal process.
What is also documented is that Marshall did not immediately collapse, did not cry out, did not appear to anyone watching to have been attacked in any conventional sense.
He felt something.
He registered it as strange.
He looked around.
The bodyguards were already moving toward the two agents.
The agents ran on foot in an Aman street in civilian clothes, having just administered a compound to a man who was now standing and watching them flee.
They ran.
The bodyguards pursued.
There was a physical confrontation.
Both Kedon agents were subdued and held.
Jordanian police arrived and took them into custody.
The four support team members read what had happened and made the only calculation available to them.
They drove to the Israeli embassy in Ammon and went inside.
Within a short time, Jordanian security forces had surrounded the building.
Israel now had two agents in Jordanian police custody and four more under siege in its own embassy.
It had a man who had been administered of poison walking into his office without knowing what was happening to him.
and it had a prime minister in Tel Aviv who was about to receive a report that would require him to make decisions in real time with no good options.
What happened inside that embassy in the hours between the arrests and the moment Marshall’s condition began to visibly deteriorate is not fully in the public record.
What is documented is that communications went out to Tel Aviv, that the situation was reported up the chain, and that the question of the antidote, still in Israel, immediately became the operational variable around which everything else would turn.
Marshall walked into his office.
He began his day.
The headache came later in the afternoon, then the nausea.
By evening, he was vomiting.
His condition accelerated.
severe headache giving way to acute respiratory distress.
He was rushed to King Hussein Medical Center.
Doctors placed him on mechanical ventilation.
His condition, according to subsequent reporting, was consistent with an opioid overdose, which was medically accurate in the sense that a synthetic opioid derivative was exactly what was killing him.
But the doctors did not know what compound they were treating because no one had told them.
because the only people who knew were either in a Jordanian police station or behind the walls of the Israeli embassy.
The Jordanian government now had simultaneously two Israeli agents in custody, [music] four more besieged in the Israeli embassy and a prominent Palestinian political figure dying in a hospital from an unidentified cause.
King Hussein connected these facts quickly.
The demand came to Israel within hours of Marshall’s hospitalization.
Provide the antidote or the detained agents will be tried for murder if Marshall dies.
The diplomatic relations built through the Wadi Araba Treaty, the treaty Hussein had sacrificed domestic political capital to sign, would not survive Marshall’s death in a Jordanian hospital from an Israeli poison.
Netanyahu received this demand and did not immediately comply.
What he was weighing during those hours can only be partially reconstructed from the accounts that exist.
The logic of refusal was not irrational, even under the circumstances, providing the antidote was an unambiguous confession.
The moment a vial of the specific compound’s antidote arrived in Ammon from Tel Aviv and was administered to Michelle, Israel had proven not only that it had conducted the operation, but that it had engineered the specific weapon.
deniability already severely damaged by the arrests was finished completely.
Whatever diplomatic cost the exposure had already created would be compounded by the confirmation.
The logic of refusal was hold, deny, let events develop and see whether Jordan could be managed without the confession.
What that logic required was that Marshall die before the political cost of his dying exceeded the political cost of confirming Israeli responsibility.
And that calculation was already failing because King Hussein had called President Clinton and Clinton was applying pressure.
And the United States was not going to watch a Jordanian Israeli diplomatic rupture over a dead Hamas official in Aman and absorb that geopolitically without demanding that Israel resolve it.
Inside Israel, the emergency cabinet meeting assembled.
Multiple proposals were raised.
Multiple proposals were dismissed as inadequate to the actual shape of the problem.
The discussion, as Halib would later describe it, went in circles because the people in the room were trying to find a solution that preserved something that was already gone.
The operation’s invisibility was gone.
The deniability was gone.
The remaining question was how much additional damage would accumulate before Israel accepted that.
Ephrame Halevi was called back from Brussels.
His presence in the crisis management was itself a signal of how seriously the government read the Jordan relationship as being at risk.
Halvi had been instrumental in building that relationship.
Had been a trusted channel to Hussein for years.
The king when he heard Haliv was being brought in asked that he not be involved.
He did not want the one Israeli he genuinely trusted to be the face of the crisis that had violated that trust.
He wanted to keep that relationship separate from the damage.
Netanyahu brought Havian anyway.
In Tel Aviv, Masho’s condition was being monitored in real time.
In Ammon, the doctors treating him were managing respiratory failure without knowing what had caused it.
In the Israeli embassy, four agents waited for instructions that had not yet arrived.
And in a Jordanian police station, the two kedon agents held whatever they held, their cover identities, [music] their training, their silence, while the hours ran down on the man they had been sent to kill.
The operation had been designed around one question.
Can we kill this man without anyone knowing we did it? That question had already been answered.
The only question left was what Israel would have to pay to keep the damage from becoming irreversible and whether the price would be set by Israel or by everyone else.
24 years before the Kedon team took up positions outside Mashall’s office in Aman, a different Mosed team made a different mistake in a different Allied country and produced a different kind of catastrophe.
The mechanics were not identical.
The underlying failure was to understand what Israel should have known about operating inside friendly territory with a lethal mandate and forged Canadian passports.
You have to go to a small Norwegian city on a summer evening in 1973 and you have to start with the man who was killed there because he is the part of the story that tends to get compressed into a footnote and he should not be.
Ahmed Buchiki was 29 years old.
He was Moroccan.
He had come to Lilahhammer, a quiet city in southern Norway known mostly for winter sports and clean air because there was work there and because Norway was a place where a young man from North Africa could build something resembling an ordinary life.
He had found that life.
He worked as a waiter at a local restaurant.
He had recently married a Norwegian woman named Terrill.
She was pregnant with their first child.
He had no connection to Black September.
No connection to the Palestinian militant networks Mossad was hunting across Europe in the summer of 1973.
No connection to Ali Hassan Salame, the Black September commander whose presence in Liil Hammer, an informant, had reported to Israeli intelligence.
He was a waiter.
[music] He went to work.
He came home.
He was by every documented account precisely what he appeared [music] to be.
The context for what happened to him begins a year earlier in Munich.
On September 5th, 1972, Palestinian militants from the Black September organization took 11 members of the Israeli Olympic team hostage in the Olympic Village.
The crisis lasted through the night and into the following day.
By the time it ended at the Fenfeld Brook Air Base, all 11 Israeli athletes were dead along with a West German police officer.
The images went around the world.
The Israeli government under Prime Minister Golden Mayor authorized what became known internally as Operation Wrath of God, a sustained campaign to track and kill the individuals responsible for planning and executing the Munich massacre.
Mosed assembled what it called hit teams and deployed them across Europe.
The operation ran for years.
Real perpetrators were killed.
Real intelligence was developed.
And then in the summer of 1973, a team received information that Ali Hassan Salame, one of the senior Black September figures linked to Munich, was in Lillahhammer, Norway.
The identification was wrong.
The mechanics of how the wrong man came to be identified as Salameé have never been fully declassified.
What is documented is that the MSAD team in Lil Hammer acted on the identification without resolving the doubt and there was doubt or there should have been because Salame was a known figure with a known appearance and the man in Lillah looked nothing like him to anyone who had reliable information about Salam’s physical description.
On the evening of July 21st, 1973, two MSAD agents shot Ahmed Buchiki multiple times on a street in Lihammer.
Terril was with him.
She watched it happen.
He died at the scene.
Norwegian police responded quickly.
The country was not at war.
It had no tradition of accommodating covert operations on its territory.
It had laws.
Six Mossad agents were arrested within hours.
The operation and Israel’s identity became fully and publicly known.
There was no exfiltration, no managed retreat, [music] no quiet diplomatic resolution.
There was an investigation, a trial, convictions.
Abraham Gamer and Sylvia Raphael, two of the arrested agents, were convicted [music] of premeditated murder and sentenced to 5 and 1/2 years in Norwegian prison.
They served less than 2 years before being released, but the trial happened in public on the record with Israeli intelligence methodology exposed in open court.
Israel paid financial compensation to Terrell Buiki and her child.
The compensation was real and inadequate, and both of those things are true simultaneously.
Canada was angered.
The Lihammer agents had used forged Canadian passports.
Ottawa had not been consulted.
Canadian documents had been used in a foreign country to facilitate a killing, the wrong killing.
And the Canadian government was left to explain to its own citizens and its allies why its passports were appearing in a Norwegian murder trial involving Israeli intelligence officers.
Now hold that sequence and place it against Ammon 1997.
Forged Canadian passports, both operations.
Lethal mandate on the soil of a country with no hostile relationship with Israel.
Both operations.
Agents caught in the field.
Both operations.
Allied government furious at the use of its documents without consent.
Both operations.
The differences matter, but so does the repetition.
After Liilahhammer, Mossad conducted reviews.
Procedures were examined.
The targeting verification process was scrutinized.
How had the identification been made? What checks had failed? What institutional pressures had pushed the team toward action when doubt existed? The stated lesson was about target verification.
Ensure the identification is correct before the trigger is pulled.
That lesson was real and it was learned in the narrow sense.
The 1997 Ammon operation targeted the right man.
There was no identification error.
Mashall was exactly who Mos had believed him to be.
But Lilahhammer’s deeper lesson, the one that the verification focus may have displaced, was not about identification.
It was about what happens to an intelligence service when it operates inside an allied country with a lethal capability and the operation becomes visible.
It was about the absence of a failure doctrine.
>> [music] >> It was about the fact that an operation planned entirely around success has no architecture for the moment success does not arrive.
In Liilhammer, the operation failed completely.
The wrong man was killed.
The agents were caught.
The damage was reputational, diplomatic, legal.
But the damage was bounded.
There was no leverage.
No one was alive whose death Israel had caused and whose survival Israel could now be compelled to ensure.
The political cost was paid in humiliation and compensation.
It was finite.
Aman would produce something Liil Hammer had not.
A dying man and an antidote.
That a symmetry, a target alive and deteriorating with the cure in the possession of the party that had administered the poison created a form of leverage that Lihammer had not generated and that Mad’s post Lilhammer doctrine does not appear to have specifically addressed.
This matters for what follows in Aman because it explains why King Hussein’s position was structurally stronger than Norway’s had been in 1973.
Norway had arrested Mossad agents and tried them in court.
That was damaging.
But Israel could absorb it, deny full institutional responsibility, express regret, pay compensation, wait for it to pass.
Hussein had something Norway had never possessed.
a man in a hospital whose life Israel could extend or withhold and two agents in a jail cell whose fate Hussein controlled.
The leverage ran in both directions simultaneously.
That is a different negotiating position.
There is one more thread connecting Lihammer to Ammon that the public record makes visible only in retrospect.
Abraham Gamer and Sylvia Raphael, convicted in the Lilahhammer trial, returned to Israel after their early release from Norwegian prison.
They eventually married each other.
They built lives.
The operational and personal consequences of Liilhammer were absorbed individually over time [music] in the way that people absorb experiences that mark them permanently but do not define their entire existence.
The two kedon agents arrested in Aman in September 1997 would eventually return to Israel too.
Their names remain classified.
What they carried out of Jordan beyond the physical fact of their release is not in the public record.
But between Liilhammer in 1973 and Ammon in 1997, 24 years passed.
And somewhere in those 24 years, in the institutional memory of an organization that conducts lethal operations in foreign countries with forged Canadian passports, the specific lesson that needed to survive, not just verify the target, but wargame the
failure, did not hold.
The Ammon team had arrived in Jordan a week before the operation.
They had done everything right by the standards of the success scenario.
They had done almost nothing to prepare for what was already in motion on the morning of September 25th.
The compound that Mosed chose for the Marshall operation was not selected arbitrarily.
It was selected because it solved a specific problem that conventional assassination methods could not solve and understanding the elegance of that solution is necessary to understanding why its failure was so complete.
The problem was this.
Marshall was living in Jordan.
Jordan was an ally.
An ally with a peace treaty 3 years old.
A king who had taken political risks to sign [music] it.
And a population that was significantly Palestinian and already watching Israel with deep suspicion.
A gunshot was impossible.
An explosive was impossible.
Any method that left a visible wound, a crime scene, a moment of obvious violence was impossible.
Not because it would fail to kill Mashel, but because it would produce exactly the kind of diplomatic rupture the operation was designed to avoid.
The kill had to look like something other than a kill.
Synthetic opioid compounds had been used in targeted operations before 1997, [music] though the specific operational history Mosed drew on is not in the public record.
What is documented is the compound’s functional profile levantel or the derivative identified as being used in this operation was engineered for transdermal and mucosal absorption.
Contact with soft tissue was sufficient for uptake into the bloodstream.
The compound then worked systemically suppressing the central nervous system with increasing severity over a period of hours.
The progression was mild disorientation, headache, nausea, vomiting, escalating to respiratory depression, and ultimately respiratory failure.
The timeline of that progression was the weapon’s core feature.
It was not designed to incapacitate quickly.
It was designed to allow a gap, several hours in the Marshall scenario, between contact and the onset of symptoms serious enough to require hospitalization.
During that gap, the subject would function.
He would walk into his office.
He would take meetings, make calls, eat lunch.
He would attribute the growing discomfort to fatigue or illness.
And critically, his bodyguards and colleagues would attribute it the same way because nothing visible had happened to him.
There was no wound.
There was no moment to point to.
By the time he was in a hospital, the agents would be in a vehicle, then an airport, then a flight to somewhere with no extradition complications.
The gap between cause and effect was the operation’s alibi.
The compound’s second engineered feature was its behavior at autopsy.
It was designed to metabolize in a way that left no clear forensic signature, or at least no signature that a standard forensic toxicology screen would catch without knowing specifically what to look for.
This was not an accidental property.
It was a design specification.
Mossad’s technical division had produced something that would kill without leaving the kind of evidence that could connect a death to a state actor.
The aerosol delivery device completed the method.
Small, concealable, capable of [music] discharging the compound in a directed burst at close range.
The ear was selected as the delivery point for specific physiological reasons.
The mucous membranes of the ear canal offer rapid absorption directly into the bloodstream faster than skin contact alone.
The discharge was quiet in open air with ambient street noise.
It would not be audible to anyone more than a few feet away.
The shaken soda can was not decorative misdirection.
It was acoustic cover for the one moment of sound the device produced.
What Mosite had built, taken as a complete system, was impressive in its specificity.
Every element addressed a real constraint.
The compound addressed the forensic problem.
The delivery method addressed the operational visibility problem.
The timeline addressed the Xfiltration problem.
The cover story.
Tourists brushing past a man on a street [music] addressed the witness problem.
The forged passports addressed the identity problem.
The antidote addressed the personnel safety problem.
It did not address anything else.
And here is the reframe.
The operation was not undone by bad intelligence or by a mechanical failure or by an informant inside the planning structure.
It was undone by a logical error that was present from the beginning.
An error so embedded in the operation’s design that it was invisible precisely because the design was otherwise so complete.
The operation was built to be deniable.
Every element served deniability.
The compound left no forensic trace.
The agents carried no obvious weapons.
[music] The passports bore no Israeli identity.
The cover story left no scene.
The entire architecture was constructed around the premise that nothing could be proven.
And that premise was sound as far as it went.
What the architecture did not account for was the antidote.
The antidote was not a security flaw introduced by careless planning.
It was a safety provision introduced by responsible planning.
In any operation involving an engineered chemical compound, the team handling it requires a countermeasure in the event of accidental exposure.
Carrying the antidote was correct procedure.
It was what a professional intelligence service does when it deploys a lethal chemical in the field.
But the antidote’s existence created a structural vulnerability that the operation’s deniability framework never resolved.
The antidote was proof, not circumstantial proof, not intelligence that could be disputed or declassified, documents that could be recontextualized, physical, chemical, biological proof.
If Mashel survived long enough to be identified as poisoned, and if the specific compound was identified, and if it became known that Israel held the specific antidote to that specific compound, the chain of inference required no interpretation, it was complete.
The operation was designed to be invisible.
The antidote made invisibility and failure mutually exclusive.
In a successful operation, the antidote was never produced, never requested, never mentioned.
Marshall died.
No one knew what compound had been used.
The antidote remained in a laboratory in Tel Aviv and was never relevant.
in a failed operation.
Specifically, this kind of failed operation, where the target was poisoned but alive, where agents were in custody, where a Jordanian hospital was treating a man for an unidentified toxin, the antidote stopped being a safety
provision and became a confession.
This is the thesis of the entire operation’s failure.
It was designed around a scenario in which the weapon and the evidence were the same object.
As long as that object was never produced, deniability held.
The moment circumstances required its production, deniability was not merely compromised, it was inverted.
Israel would have to produce the proof of its own act in order to survive the consequences of that act.
No intelligence service builds an operation intending to produce that proof.
But the Ammon operation made the production of that proof inevitable.
The moment the partial failure occurred, the agents were in custody, the weapon had been used, the target was dying.
The only path out of the crisis that did not result in dead agents, a ruptured peace [music] treaty, and the full political cost of Marshall’s death on Jordanian soil was the antidote.
And the antidote said, “Israel [music] did this.
” In Tel Aviv, the emergency cabinet was assembling.
[music] King Hussein had called Clinton.
The clock in the hospital was running.
The four agents in the Israeli embassy were waiting.
The two agents in Jordanian custody were waiting.
And Mashall on mechanical ventilation was somewhere between the operation that had been designed to kill him invisibly and the diplomatic crisis that was going to make him famous.
Netanyahu had the antidote.
He had not yet decided to use it.
The decision he was avoiding was not a decision about whether to save Marshall’s life.
It was a decision about whether to confess and what that confession produced under duress in the capital of an allied country under pressure from the president of the United States would cost Israel in every direction simultaneously.
He had hours at most before the decision was made for him.
Netanyahu’s first response to the crisis was not to provide the antidote.
It was to send Danny Yatum to Aman.
This was the false start.
The reasoning was not irrational.
Yatam was the MASA director.
The operation was his and a direct channel from the intelligence chief to Jordanian counterparts might contain the situation before it fully exploded.
If the two arrested agents could be recovered quietly, if the Jordanian security services could be persuaded that this was a matter best handled without public exposure, the operation’s damage might be bounded.
Yatam flew to Aman with Netanyahu’s consent.
He arrived carrying no antidote.
He arrived carrying the implicit message that Israel wanted its people back and was prepared to discuss terms.
What he was not prepared to do, what Tel Aviv had not yet authorized was confirm that Israel had poisoned Marshall and that Israel held the cure.
King Hussein did not receive Yatam warmly.
The Jordanians read of the situation was already complete.
They had two men with forged Canadian passports in custody.
They had a dying Palestinian political leader in a hospital.
They had four more Israelis inside the Israeli embassy [music] surrounded by Jordanian security forces.
The geometry of the situation did not require elaboration.
What Hussein wanted was not a conversation about terms.
He wanted the antidote.
He wanted it acknowledged.
He wanted it delivered.
Yatam’s arrival without the antidote read in Ammon as Israel still trying to manage a situation that was no longer manageable.
The Jordanians indignation as Halvie would later describe it deepened.
Hussein escalated.
The message back to Netanyahu was direct.
Provide the antidote or the agents face murder charges and if Mashall dies they will be hanged.
Yatum returned to Tel Aviv without resolving anything.
In the hospital, Mashall’s condition was worsening by the hour.
The doctors treating him were working without the one piece of information that would have made their treatment precise.
They had identified the symptomatic profile, opioid toxiddrome, respiratory depression, the pattern of a central nervous system under suppression from a chemical agent.
They were managing the symptoms.
Mechanical ventilation was keeping him breathing.
But without knowing the specific compound, they were treating the effect rather than the cause.
The antidote, sitting in a laboratory in Tel Aviv, was calibrated to the specific molecular structure of the compound that was killing him.
Without it, the doctors were improvising against something they could not name.
The window in which the antidote would be effective was not infinite.
Back in Jerusalem, the cabinet meeting was going in circles.
Options were proposed, options were rejected.
A narrative was being constructed in real time in which Israel might still find a way to limit the damage.
Some formulation, some channel, some framing of events that would let Israel acknowledge something without acknowledging everything.
Clinton’s call ended that the president of the United States had spoken to King Hussein.
He had heard the Jordanian account.
He had registered that the Wadi Araba peace treaty, a cornerstone of American Middle East policy, a diplomatic achievement the United States had invested in significantly, was at genuine risk of collapse because Israeli intelligence had poisoned a man in Ammon without telling anyone and was now stalling on the antidote.
Clinton’s message to Netanyahu was not a suggestion.
Netanyahu held not because he believed he could still win the situation.
The cabinet meeting made clear there was nothing to win.
He held because the moment he moved, the last remaining question whether Israel had done this would be permanently answered.
Providing the antidote was not a concession.
It was a confession.
And Netanyahu understood, as every leader in that position understands, that the sequence of confession and concession would be set by the people with leverage, not by Israel.
For a brief period, there was a moment that felt internally like a possible off-ramp.
Word moved through intermediaries that Jordan might be satisfied with a less complete Israeli acknowledgement.
Something that addressed Marshall’s medical situation without a full formal confirmation of Israeli responsibility.
A formulation in which the antidote arrived through a channel that preserved some ambiguity about who had supplied it and why.
This was the false release moment.
The room in Jerusalem read it as a potential path.
It was not a path.
It was Jordan testing whether Israel understood yet how completely the situation had turned.
The test failed immediately.
The Jordanian position did not soften.
The ambiguity formulation was not accepted.
Hussein’s office made clear that the terms had not changed.
The antidote acknowledged delivered directly with Israeli confirmation of what it was and what it was for.
The false release cost time, more hours.
In the hospital, Marshall’s condition did not wait.
Ephrame Halvie had been called back from Brussels.
He arrived in Jerusalem and went into the cabinet meeting.
The discussion he entered was one that had already exhausted its alternatives.
What Halvie added was not a new option.
It was a frame [music] for accepting the inevitable in a way that preserved something rather than nothing.
His argument was this.
Israel was going to pay.
The question was not whether to pay, but what the payment bought.
If the payment was structured as Israeli compensation, as Israel buying its way out of the crisis, it would read as weakness in every Arab capital and confirm that Israel could be coerced.
If it was structured as Jordanian strength, Hussein extracting a price that Israel was compelled to meet.
Hussein could defend it domestically.
The peace treaty could survive and the damage, though severe, had a ceiling.
The distinction required Israel to accept a larger concession than it wanted, framed in a way that served Hussein’s domestic position rather than Israel’s.
It required Netanyahu to let Hussein set the terms publicly.
Netanyahu was not ready.
He was still not ready when Clinton’s pressure arrived a second time, harder.
He was still not ready when the hospital sent an update on Marshall’s condition that made clear the window for effective antidote administration was narrowing toward closure.
Then he called Halib.
The accounts of that call are consistent across sources.
Netanyahu’s voice was the physical register of a man who had been inside an impossible situation for too many hours without sleep or resolution.
He said two words in English.
do it.
He hung up.
Halvie left for Ammon.
The antidote left Tel Aviv on a light aircraft.
It arrived in Jordan.
It was delivered to royal physician Dr.
Sammy Rababa.
He administered it to Mashall at King Hussein Medical Center.
The compound in Mashel’s system, the Leventinyl derivative, the engineered silence that was supposed to make him simply cease began to be reversed.
He had hours to spare.
Not many.
In the Israeli embassy, the four besieged agents waited.
In the Jordanian police station, the two Kaidon agents waited.
The operation’s cover identities, Barry Beads, Sha Kendall, the Canadian tourists who had come to Ammon, for reasons no one had fully explained, [music] were already as useful as the passports they were printed on, which is to say not useful at all.
The antidote’s arrival in Ammon was the moment the last door closed.
There was no longer any question of Israeli deniability.
[music] There was no longer any question of whether the operation had happened, whether Israel had authorized it, whether the compound was an Israeli weapon, or whether the man in the hospital had been deliberately targeted.
The antidote answered all of those questions simultaneously.
It answered them in a Jordanian hospital, in the presence of Jordanian physicians, [music] in a country whose king was on the phone with the president of the United States.
The operation designed to leave no evidence had just produced the most complete evidence possible.
Netanyahu would fly to Ammon 4 days later on September 29th to apologize personally to King Hussein.
He would not be received by the king.
He would be met by Crown Prince Hassan.
The message in that detail was not subtle.
It was not meant to be.
The apology tour was beginning.
The reckoning had not yet started.
The price was set by everyone except Israel.
That is the first thing to understand about the fallout.
The sequence of concessions that followed the antidotes delivery was not negotiated in any conventional sense.
It was dictated by Hussein’s leverage, by Clinton’s pressure, by the arithmetic of eight Israeli agents in Jordanian hands, and a peace treaty that could not survive Marshall’s death.
Israel’s role in the negotiation was to accept terms in an order and at a pace determined by the other side.
Hiv reached Ammon and worked through intermediaries before Hussein agreed to meet him directly.
The king had specifically asked that Halevi not be involved.
He did not want the relationship he trusted contaminated by the crisis that had violated it.
He relented because the deadlock required someone Hussein could speak to without the conversation becoming purely transactional.
The meeting was tense.
Hussein was agitated in a way that Halavi who had known him for years read as genuine rather than performed.
The indignation was not diplomatic positioning.
It was personal.
Israel had treated Jordan as a country where it could act without consequence.
The man across the table from Hlevy had staked his political credibility [music] and in some measure his physical safety on the argument that peace with Israel was worth the domestic cost.
The Aman operation had been conducted as though that credibility did not exist or did not matter.
Halivvi raised the release of Shik Ahmed Yasan.
Hussein agreed immediately.
Then they reached the harder problem, the agents.
The two Kedon agents in Jordanian police custody required a formal resolution.
Charges dropped, transfer arranged, departure from Jordan.
The four in the embassy required Jordanian security forces to stand down and allow them to leave.
Hussein needed to be able to explain each of those decisions to his security services, to his cabinet, to a Jordanian public that had watched Israeli agents caught in the streets of the capital.
The framework Halevi had brought from the Jerusalem cabinet meeting that this must appear as Jordanian strength, not Israeli management, was the only architecture that made those explanations possible.
It worked in the narrow sense that it produced an agreement.
It did not work in the sense of limiting the cost.
Sheikh Ahmed Yasan was released from an Israeli prison in early October 1997.
He was 60 years old, quadriplegic, nearly blind, serving a life sentence.
He had been in Israeli custody since 1989.
His release was the price of recovering two Kaidon agents from a Jordanian jail.
Approximately 70 additional Palestinian prisoners were released alongside him.
Yasin returned to Gaza.
He was received as a figure of historic stature, the founder of Hamas, freed by Israel’s own failure.
The terms of his release included a condition that he refrain from calling for attacks against Israel.
[music] He violated that condition within weeks.
He resumed his leadership of Hamas, coordinated its activities, and continued to do so for 7 years.
In March 2004, an Israeli Apache helicopter fired three missiles at him outside a Gaza mosque as he was leaving morning prayers.
He was killed.
His two bodyguards were killed.
Nine bystanders were killed.
The man Israel had been compelled to free in 1997 to recover its agents was killed by Israel in 2004 [music] after seven additional years of leading the organization whose political chairman Mossad had tried to decapitate in Aman.
The operation intended to weaken Hamas’s external leadership had instead liberated its founder.
Netanyahu flew to Aman on September 29th.
He was not received by King Hussein.
Crown Prince Hassan met him instead.
The signal required no interpretation.
Hussein would not give Netanyahu the image of a reconciliation between equals.
The humiliation was calibrated and deliberate and entirely justified by what had happened.
Canada recalled its ambassador to Israel.
Foreign Affairs Minister Lloyd Axworthy proposed the recall.
Prime Minister Jean Cretian sanctioned it.
A temporary freeze on intelligence sharing followed.
In McClean’s magazine, a senior Ottawa adviser described the government as mightily pissed off.
Language that in Canadian diplomatic register signaled genuine institutional anger rather than performed outrage.
The Canadian passports had been used without consent.
This was not the first time.
The patients that had absorbed previous instances had reached its limit.
The Sahanover Commission reported in the months that followed.
Its findings held Yatum accountable for the operational shortcomings.
On February 24th, 1998, Danny Yatum resigned as Mossad director.
In his resignation letter, he stated precisely and without ambiguity [music] that he categorically disagreed with the commission’s conclusions.
He resigned anyway.
Institutional responsibility in his [music] framing required acceptance of consequence regardless of personal disagreement with the verdict.
The first chapter of his memoir covers the operation.
He discussed it in interviews for the rest of his public life.
He was not broken by it, he was marked by it, which is a different thing.
Ephraim Halevy, the man who had negotiated the crisis into something survivable, was appointed Messiah director to replace him.
He served from 1998 to 2002.
Then he headed the Israeli National Security Council.
He discussed the Ammon operation publicly on its 25th anniversary [music] in 2022 with the specificity of someone who had spent decades thinking about what it meant.
He did not describe what he had done as heroic.
He described it as damage control conducted inside a failure that should have been anticipated.
King Hussein died in February 1999, 17 months after the crisis.
Cancer.
He had been diagnosed and treated intermittently for years.
He spent his final months managing the Israeli Palestinian peace process, the handover to his son Abdullah, and the accumulated weight of a reign conducted across four decades of regional upheaval.
He did not see Hamas expelled from Jordan that happened in August and September 1999.
under Abdullah who moved against the organization’s presence in the kingdom for exactly the stability reasons Israel had cited in 1997.
The diplomatic pressure that had been available to Israel all along and that it had chosen not to use worked.
It worked 23 months after the attempt to bypass it with the poison.
Marshall recovered.
He returned to work.
He led Hamas’s political bureau until 2017, 20 years after the attempt on his life.
When Israel killed Yasan in 2004, and Yasan’s successor, Abdelaziz al-rantisi the following month, the man Mossad had tried to kill in a street in Aman became the recognized head of the organization.
He was, in his own words from a 2008 interview, more resolute after the attempt than before it, more certain that his survival was meaningful, more committed to continuing.
The operation had not weakened his position in Hamas.
It had consecrated it.
As of late 2024, following the deaths of Ismael Han and Yaha Sininoir, Mashall is again the acting leader of Hamas.
The last major figure of his generation still standing.
[music] The two Kedan agents returned to Israel.
Their real names remain classified.
They faced no judicial consequences.
What they carried out of Jordan beyond the physical fact of their release is not in the public record.
Now pay off Lil Hammer.
In 1973, six MSAD agents were arrested in Norway after killing the wrong man with forged Canadian passports.
Israel paid compensation.
The agent served abbreviated sentences and came home.
The institutional lesson as it was recorded and transmitted was about target verification.
Confirm the identification before you act.
In 1997, Mossad confirmed the identification.
Mashal was exactly who they believed him to be.
The Canadian passports were used again.
The Allied country was Jordan instead of Norway.
The agents were caught again.
The lesson that did not transmit across 24 years, across internal reviews, across a generation of Mossad leadership was not about who you target.
It was about what you do when the operation fails inside a country that has leverage over you.
It was about what happens when your weapon and your evidence are the same object.
It was about the specific catastrophe produced when an operation designed to be invisible becomes through partial failure the most visible thing your government has done all year.
That lesson was not written down in a form that held.
13 years after Aman in January 2010, Mossad operatives traveling under forged Canadian identities were caught on surveillance cameras in Dubai after the assassination of Hamas military commander Mahmud al-Mabu.
Different operation, different decade, same document.
An operation designed to be invisible forced Israel to act in the most visible way possible, producing publicly the antidote that proved what it had done.
That is the shape of the failure.
Not the poisoning, not the arrests, [music] the antidote.
The safeguard built into the operation to protect its own people became the instrument of its complete exposure.
The more carefully the operation was planned, the more thoroughly it destroyed itself when the plan broke.
Intelligence doctrine calls this a failure of secondord thinking.
The inability to model what your own safeguards look like from the outside when everything goes wrong.
The ammon operation is the clearest example of that failure in the modern record.
It is studied because it is instructive.
It is instructive because it is not unusual.
The operations most likely to produce this kind of catastrophe are not the poorly planned ones.
They are the ones that are planned carefully for one outcome.
with nothing left over for any other.
If this is the kind of documented, verified intelligence history you come here for, the record as it actually happened, not as anyone wished it had, subscribe to Hidden Ops.
New episodes drop regularly.
No invented stories, no dramatized conclusions, just the operations, the decisions, and what they actually
Disclaimer : This content may be created by AI for entertainment purposes. Any resemblance to real persons, events, or places is coincidental.
