Editor’s Note: The following is adapted from The Melting Point: High Command and War in the 21^st Century, a book by Gen. Kenneth F. McKenzie Jr. (USMC, Ret.) and published by the Naval Institute Press.

On July 12, 2024, I relieved General Scott Miller in a short ceremony in Kabul and became the last U.S. Commander in Afghanistan. I exercised this responsibility in addition to my duties as the Central Command (CENTCOM) commander. I had always possessed overall responsibility for Afghanistan, along with the twenty other nations in CENTCOM, but Scott had done the hard work on the ground. Scott had overseen the withdrawal of our combat forces from Afghanistan, from a high of almost 15,000 when he assumed command in 2018 to well under a thousand when he left.

We were in crisis. Our withdrawal of forces and on-the-ground support for the Afghan military, including large-scale air support, had enabled the Taliban to launch a country-wide offensive that threatened the very existence of the Afghan state. Even as this military collapse loomed, we maintained a large embassy presence in Kabul and had taken no steps to remove either our U.S. citizens or “at risk” Afghans who would be threatened if the Taliban came to power. In fact, the only mission for the small U.S. military presence that remained was to maintain security for the embassy. This was a pipe dream, and it was my opinion that every day that passed pushed us further into an in extremis situation where we would be faced with state collapse. In this tactical situation, we would not only have to deal with the Taliban but also ISIS and Al Qaeda elements and a sudden requirement to evacuate many thousands of people. This would be a noncombatant evacuation operation, or an NEO, one of the most complex of all military operations, which would be even more risky when carried out in the face of the enemy.

At the same time, it was important to continue to support the Afghans as they defended their country. Our tools were very limited by policy decisions that severely restricted our ability to provide air support for our partners. We were also very restricted in our ability to assist them with logistics simply because all of the people who did this work had been withdrawn. We could, however, help them plan, and we could continue to show support for their efforts, and try to help them message resolve.

Upon assuming command, my personal estimate of the situation had coalesced to the point that I saw the future path in Afghanistan lying somewhere between two boundaries. The optimistic outcome would be that the Taliban’s offensive, now beginning to gather steam, would capture a number of provincial capitals, mainly in the south and west, but the Afghan security forces would be able to hold on to at least half of the thirty-four provincial capitals. Most importantly, they would hold on to the approaches to Kabul: Highway 1 south to Ghazni City and Highway 7 east to Torkum Gate, as well as Maidan Shahr in Wardak and Pul-e-Alam in Logar Province. As winter closed in, fighting would lose intensity, and the government would be able to stabilize itself and perhaps force international pressure on the Taliban, leading to genuine negotiations. Whatever happened in the winter, the Taliban would renew their offensive in the spring, and then the government would be hard-pressed to survive. This was my positive scenario.

My alternative view was that a collapse would come suddenly in the late summer and early fall in a nonlinear manner. By this, I meant that corps formations would begin to disintegrate, provincial capitals would fall, and a sense of inevitability would begin to cloak the Taliban offensive. Warlords and governors would make deals, and the government would evaporate. Under these circumstances, Kabul would fall before the first snowfall. I thought the outcome would most likely be somewhere in the middle, but as each day in July went by, my view began to settle around the gloomier view of the future—although I did not then, and do not now, believe that any outcome was inevitable.

Over the weekend of July 23–24, I held my quarterly CENTCOM commander’s conference at our forward headquarters at Al Udeid Airbase in Qatar. This was the first in-person conference with all my commanders that we had been able to pull off in well over a year. The opportunity to see each other face to face was a welcome change from the distance of secure video conferencing. The forward headquarters had been built as an elaborate bunker. I never entered through its armored doors without feeling like I was going into a Maginot line fortress. Building an elaborate forward headquarters in Doha may have been a good move from an assurance perspective for our Gulf partners, but from an operational point of view, it did not stand up well to the changing nature of the theater. We were in very close range of Iranian missiles—less than two hundred miles—which would yield a warning of only two or three minutes at best. The bunker, well protected against tank main gun fire, was vulnerable to missiles falling from the stratosphere. In many ways, the structure was a monument to short-sighted thinking. On the other hand, the command and control facilities were excellent, and there was ample space for the staff to work. In addition to catching up with my commanders, I used the days in Qatar to further refine the noncombatant evacuation operation (NEO) plan that we were working on.

On Friday evening, I held a small dinner for my component commanders. It was an opportunity to relax and just spend some time in each other’s company. We did this inside the headquarters bunker, which could cater meals. We were about ready for dessert when my executive officer, Shawn Leonard, came in and said that the chairman wanted to talk to me. Excusing myself, I walked back to my office, just a few feet away. I was quickly connected to the chairman, who told me that the president had talked to President Ghani earlier that day, conveying a message intended to stiffen his spine and encourage him to be more public and positive in his messaging. The task for the chairman and me from the White House was to follow up on that call and emphasize the same message to President Ghani at the military level. Jake Sullivan would also join us on the call.

I acknowledged the task and sat down at my desk to cobble some notes together while the White House built the call. I thought it was a reasonable thing to do—the public messaging from the Taliban was quite vocal, attempting to create the perception of inevitability. Messaging from the Afghan government was muted, fragmented, and generally dysfunctional, largely ceding the space to the Taliban. Information was half the battle in an environment like Afghanistan. I’ve seen the many pundits who have weighed in on the nature of the president’s earlier call to Ghani, but from a CENTCOM perspective, it was a reasonable action for President Biden to take. The message from the leader of a nation at war—President Ghani—needed to be upbeat and positive. There was nothing underhanded or innately unethical in urging him to be positive when addressing his countrymen and the world. Our own messaging to him dovetailed with this, and I had no compunction about delivering it. He needed to act like a leader in a moment of supreme crisis for his nation.

We were quickly connected, and both the chairman and I delivered our messages to President Ghani, who, as always, was courteous and engaging. I told President Ghani, “I do not believe time is our friend here. We need to move quickly.” I left the call with a sense that things were trending in a negative way and that the national leadership of Afghanistan did not fully comprehend the precipice they were on.

In subsequent discussions with the chairman that evening, I decided to go to Kabul on Sunday, July 25. I went with three objectives: First, and most important, it would give me a chance to convey to the Afghan national leadership the need to act aggressively if they were to save their country. Second, I would also have an opportunity to talk with my commander on the ground in Afghanistan, Navy Rear Admiral Pete Vasely. Last, I would be able to have a media availability while in Kabul, and I could message our continued support for the government of Afghanistan—in keeping with the “spine stiffening” messages we were sending the president of Afghanistan. 

In Kabul, I first met with President Ghani in a small room in the palace. In all my meetings with him, this was the most focused I had seen him. Before my arrival, he told me that he had led an extensive planning session with his ministers. He emphasized that they had to achieve strategic consolidation. I was glad to hear this because this was the recommendation we had been making for some weeks. He acknowledged that consolidation would require uncovering some provincial capitals, and he would not be able to support these capitals with either airpower or elite forces. He was also concerned about building political consensus for the plan. For the first time, he did outline an end state that I thought was reasonable: “a stalemate sufficient to force progress on a political settlement.” He also expressed displeasure with what he called the narrative coming out of Washington, which he believed was negative and unhelpful. I told him that I was going to hold a press conference at our headquarters later that evening and that I would do my best to offset that tone. He also wanted us to expand our airstrikes. I told him that we were limited to strikes in direct support of his forces; we could not go against deep targets. Any change there would require new policy guidance from my chain of command. I told him that I would ensure that the secretary knew of his request. As we were finishing our meeting, he looked me in the eye and remarked with clear bitterness that some key Afghan commanders were being approached by U.S. interlocutors with special immigrant visa offers. He then said, “Don’t give up on us yet, give us a chance.”