Lisa Ling began her military career in the early 1990s as a medic and nurse, later gaining recognition for her expertise in information systems. Encouraged to shift into combat communications, she went on to work on the operations, maintenance, and security of networked military technology. As the U.S. military expanded its Intelligence, Surveillance, and Reconnaissance (ISR) capabilities, her Combat Communications Squadron was absorbed into the drone program and relocated to Beale Air Force Base in California.
Over the course of her service, Ling was stationed at key military sites, including the Distributed Common Ground System (DCGS) headquarters at Joint Base Langley-Eustis in Virginia, an Air National Guard unit in Kansas, and several overseas deployments.
She ended her active-duty career at Beale and later traveled to Afghanistan to witness firsthand the impact of the systems she once participated in.
In the years since, Ling has become a whistleblower and researcher, critically examining the convergence of AI, surveillance, and automated warfare. Together with fellow whistleblower Cian Westmoreland, she coined the term Kill Cloud, a term first introduced in the anthology Whistleblowing for Change edited by Tatiana Bazzichelli (transcript Verlag, 2021)—to describe the sprawling, networked infrastructure enabling data-driven killing.
This interview was conducted in partnership with the research programme Investigating the Kill Cloud: Information Warfare, Autonomous Weapons & AI (2023–2024), developed by Disruption Network Lab to bring together whistleblowers, researchers, and artists to interrogate the future of warfare and the growing entanglement of military systems with artificial intelligence.
Walid El Houri: Can you explain the “Kill Cloud” concept and how it encapsulates the evolution of networked warfare?
Lisa Ling: The term kill cloud mostly came from Cian Westmoreland as we tried to find a way to explain what’s really happening. The concept itself isn’t laid out in a way that connects all the moving parts—because there are so many.
To understand the kill cloud, think about everything we have connected to networks today: refrigerators, TVs, phones, computers. All of these devices communicate with each other to make life “simpler.”
Technology has been helpful in many ways. But when the same systems are used to target or kill, the connected devices are no longer just printers or mice. They become drones, weapons, robot dogs, and other tools that can be disconnected, loaded with data, and deployed to kill. The technology isn’t radically new, it’s just scaled up.
Corporations like Amazon, Microsoft, and telecom companies have built the infrastructure that moves data, and the military now uses that same infrastructure, not to send emails or order a ride, but to identify and target people. So when we order an Uber, that’s data coming from the cloud. And when the military uses the same cloud-based systems to kill, it makes sense to call it the kill cloud, because that’s what it is.
WH: How has the shift from geographically confined battlefields to global, interconnected military systems changed the ethics and accountability of war? How do you think the public’s understanding of drones especially in the West obscures the broader implications of network-centric warfare?
LL: When wars began, it was person against person. You sent battalions, large armies. It all took time. Now, take something like Elon Musk’s reusable rocket: you could load it with military equipment and, later, personnel, and send them anywhere on the planet within an hour. The barrier of distance is gone.
Another major barrier used to be public outcry. When people’s sons and daughters were sent to war, families protested. But now we’re not sending as many people, we’re sending machines. That shift reduces public resistance and feeds myths: that war is now safer, that strikes are precise, that conflicts will be shorter. In reality, what we’ve seen is endless war.
Drone strikes have crossed borders into places not officially at war, and the laws of armed conflict, international humanitarian law,all once considered customary, are now eroding in favor of what I would call a techno-colonial ethic.
When we talk about smaller drones, often called the “weapon of choice of the poor,” we’re not looking at network-centric warfare. These are local, line-of-sight systems, often used to collect data in place of risking human life. In fact, when drones were first deployed, at least by the U.S., they didn’t carry weapons. They were used for overwatch. You could observe two opposing forces and maybe redirect one to avoid unnecessary death. But once weapons were added to drones, just like with any platform, their primary function shifted: from surveillance or defense to killing. That’s when defensive protective technology becomes offensive killing technology.
WH: What are the most concerning implications and pressing issues of integrating AI into military systems, such as autonomous targeting platforms? What has the Israeli war on Gaza shown in this regard for the future of warfare elsewhere?
LL: There are a lot of implications here. War planners and war makers are saying many things about this technology, and even those who oppose it still talk about “putting a human in the loop.”
Let’s say the system scans over 700,000 records and narrows that down to a subset shown to the person tasked with pressing the button to fire a weapon. If that person, the so-called “human in the loop,” like the one reportedly involved in Israel’s operations, can’t meaningfully read or process that data, does their presence actually change anything? I don’t think it does. It’s like clicking “agree” on an end user license agreement—most of us don’t read it, we just press enter. There’s this built-in trust that the computer must be right, that what it produces is inherently just. But the evidence, especially in Gaza, doesn’t support that belief.
So who’s responsible when a missile is fired? The person in the loop? The commander? The programmer who built the AI? The corporation, Microsoft or Google, that provided the system? We don’t even have a framework for how that responsibility is assigned.
In a paper I wrote, I discussed the AEGIS weapons system. Compared to newer systems, it’s relatively simple: a plane either sends a signal—“I’m friendly” or “I’m a civilian”, or it doesn’t. Even then, during the investigation into the downing of Flight 655 over Iran, both sides disagreed, and it was messy. But at least there was an audit trail to verify the information.
Do these new drone and AI systems have an audit trail? I haven’t heard much discussion about oversight in that form. Most of what I hear is still about humans in the loop.
Meredith Whitaker talks about large language models (LLMs) being used by the military. These models ingest vast amounts of data scraped from the internet, data about you, me, anyone who’s used a credit card, signed onto a network, or interacted with any cloud-connected system like Uber. Anyone could potentially be flagged as a target.
Take something as simple as frequent movement. In some cultures, that’s normal and appropriate. In the U.S., young people move around a lot. That doesn’t mean they’re terrorists, but in these systems, such behavior can become a trigger for targeting. How does that make any sense?
And there’s a real difference between someone forced into a situation by circumstance and someone actively working to harm others. But I don’t see those distinctions built into the logic of these systems.
What I do suspect is that most of the incoming data in these devices is framed positively, “this is who to target.” But is there anything in the system that says, for example, “do not target a U.S. NGO,” like in the last drone strike in Afghanistan? Clearly not. I don’t have evidence that AI was involved in that strike, but I wouldn’t be surprised if it was.
WH: How can oversight mechanisms keep up with the rapid evolution of technologies like the “Kill Cloud”? And is there really the ability to have real oversight let alone accountability within the present world system?
LL: I think treaties and similar mechanisms could help, but we’re talking about intergenerational efforts. Look at how long it took to address nuclear weapons, landmines, or cluster munitions. How many people had to die before those agreements were even considered? And even then, will such treaties be honored?
Do I see any effective oversight today, with the systems we have? Honestly, I don’t. Will international humanitarian law hold? It doesn’t seem so now. Even if one doesn’t consider what’s happening in Palestine to be genocide, there are clearly war crimes being ignored. So can anything be done when powerful countries can veto action at the UN Security Council? I don’t know. But right now, I don’t see a viable path forward.
As for governing these technologies, you’d need to navigate patent law, corporate protections, military secrecy, state secrecy. I don’t know how any group or committee would gain access to what’s actually feeding these systems.
In scientific research, there are rigorous methods for collecting and validating data, whether through double-blind studies or other structured approaches. But the way these targeting systems collect data is nothing like that. From what I’ve seen, it’s about collecting everything: buying from data brokers, gathering it yourself, ingesting it from networked systems. There’s no disciplined or transparent method behind how data is used to target buildings, weapons, or people. I just don’t see it.
WH: What role do major tech companies play in the development of military systems, and how does that affect public accountability? How can we hold such massive companies accountable?
LL: This brings up another aspect of networked warfare: propaganda. These systems have the capacity to spread information far beyond the boundaries of any one nation-state, and then there’s geo-censorship. You might think you’re seeing everything when researching online, but you’re not. For example, while I was in Berlin, I could access certain sites that became unavailable once I returned to the U.S. That really shifted my thinking.
As with many other issues, the old saying applies: Follow the money. You can see who profits from this system of war, figures like Peter Thiel, Elon Musk, Mark Andreessen. They’re venture capitalists, and that’s a major conflict of interest in the current U.S. context.
I’m not a financial expert, but it’s clear that much of Silicon Valley has lobbied Congress for expanded powers and influence.
I now see companies like Microsoft and Google as weapons contractors, no different from Raytheon or other legacy defense firms. Think of all the employees who left Google over Project Maven, that was part of a much larger weapons system. And if negotiations to stop the war in Palestine were happening at the same time as defense contracts were being signed with Microsoft and others, then what we’re looking at is a stalemate.
WH: What do you foresee as the next major challenges in regulating AI and networked systems in warfare? Do you see hope that there can be ethical use or better governance of these technologies in the future?
LL: I’d like to think there’s hope. I really would. Human beings are a hopeful lot. But right now, what seems to be missing is the political will of nation-states to prioritize governance before technological advancement.
There’s an enormous dependence on artificial intelligence, and that dependence is overstated, both in terms of accuracy and the myth that AI can be unbiased. All of this data comes from a world with a long history of bias and racism, especially in the West. So how can we expect unbiased outcomes from systems built on biased data?
Where should governance happen? Do we regulate the large language models before they’re handed over to the military? Do we regulate the data brokers and miners, what they collect, what they sell? Do we regulate credit card companies and what they share with brokers?
There used to be strong intelligence oversight in the military, but even that has changed. Now, accountability is so diffused that in a single strike, multiple entities are involved. So where does accountability lie?
I don’t think any of these questions have been clearly answered, or even seriously addressed. I hear people talking about how today’s soldiers are learning to trust AI, about human-AI teaming with robotic dogs and other systems. But what I’m not hearing is talk about oversight, governance, or human rights.
Warfare no longer looks like it used to. Your enemy no longer wears a uniform. Anyone can buy a drone off the shelf, strap a Molotov cocktail to it, and send it flying. The accessibility of this tech has created a mess.
I’m still hopeful, because there are still people who care about justice, fairness, and humanity. I believe enough of those people exist that something can change. But the ubiquity of these systems makes it harder.
Everyone can see the drones, the hardware, but what’s hidden is how these aren’t truly “unmanned” systems. They’re hypermanned, just not onboard. And because these technologies are embedded in our everyday lives, through communication systems, smart devices, and cloud platforms. They’ve become invisible and normalized.
It used to be hard and expensive to make a long-distance call. Now, it’s hard not to be connected. And that makes this fight more complex.
But if people start talking about these difficult topics, war, surveillance, militarization, and start engaging as if their own children were being sent to war, it could make a real difference. In the past, those sent to war sometimes had parents in power. Now, wars can be executed from a hometown. That’s a major shift.
The character of war has changed, but war itself hasn’t. It’s still violent. It’s still one of the ugliest things we humans do. And if we can start speaking honestly about that, at the community level, we might begin to push for change.
