The Future of AI Combat Aircraft | Behind the Wings Podcast

Neural Jetstream
2 followers
Feb 16, 2025
In this episode, we’re switching things up to bring you a special conversation with Secretary of the United States Air Force Frank Kendall.
We had the opportunity to visit the Secretary at the Pentagon to discuss AI’s evolving role in U.S. air power, collaborative combat aircraft, technology breakthroughs, and ethical concerns. It’s time to go Behind the Wings!
We're in a race for military technology superiority, and if we don't move forward,
we're going to lose. What happens if an adversary starts deploying this kind of technology sooner?
The disparity in time between when we can feel this kind of capability and when they have
something comparable is not going to be very much. Critics would say that there are concerns with the
potential of things going wrong with AI taking over. It's a very messy, ugly, complicated,
confusing endeavor. Are we ever going to get to a point where the machine is actually making
that lethal decision? Human beings are always going to have to have a role in overseeing it.
Where do you see it going over the next 10 to 20 years? We're integrating artificial intelligence
in various forms into our systems today. It will be a paradigm shift. Hi, I'm Tracy Lauderette,
Colorado's First Lady Fighter Pilot. In this podcast episode, we're excited to bring you
the full interview with Secretary of the United States Air Force Frank Kendall. We had the
opportunity to visit with the Secretary at the Pentagon to discuss AI's evolving role in the U.S.
military, collaborative combat aircraft, technological breakthroughs, and ethical
concerns. This one is going to be cool. It's time to go behind the wings. As Secretary of the Air
Force, how do you view the integration of AI as part of the strategic integration of the future?
What we call AI, artificial intelligence, is really a collection of a number of technologies.
It's not human-like intelligence, but it's a lot of capabilities that can contribute to our military
suite of things that we're trying to do. Autonomy is at the top of that list, probably. The ability
of machines to operate largely on their own, perhaps with human supervision, but not with
direct operation by humans. The suite of things that we can have aircraft do, for example,
using AI technologies is increasing constantly over time. It's gotten to a level where we can
make a major commitment to utilization of that kind of technology. You worked at the Pentagon
back in the 80s. Autonomy was something that was front of mind. How have we gotten to this place
now where AI, where artificial intelligence, is really the name of the game? It's been an
evolutionary set of technology developments with revolutionary applications. I was one of the
autonomous aircraft in particular, as well as other platforms, going back, as you said,
to the 80s when I was director of tactical warfare programs in OSD. I've watched that
technology over the last almost 40 years move to a point where we can increasingly rely on it for
some of the functions that traditionally humans have done as operators. Autonomous weapons have
really been described as a third revolution in warfare, starting with gunpowder and the nuclear
arms. Do you see AI as a paradigm shift? Is this the new way we're going to be doing business?
I think it will be a paradigm shift. I think it's happening as we speak. This is a change
that's relatively recent. A few years ago, when I was in a different position in the Pentagon,
I initiated a program called the Aerospace Innovation Initiative, which was designed to
field the next generation of fighters to develop technology for that generation,
the follow-on to the current F-35, if you will. And I worked that with the director of DARPA,
the Defense Advanced Research Projects Agency. And there was a big question mark at the time,
should we go at that time to embracing uncrewed aircraft with a full range of military functions?
We didn't do it then, but we're ready to do it now. And that's just in the last several years.
So we're seeing the integration of artificial intelligence being tested with fifth generation
fighters. And you're an Army guy at heart. So how do you see us integrating and developing AI
in conjunction with the sister services moving forward?
Well, first of all, I may be an Air Force and Space Force guy at heart. I did start out in the
Army, and I spent quite a few years there. But I'm an aerospace engineer, and I've been working
around space and aviation programs for almost my entire career. If it hadn't been for my eyes,
I would quite likely have gone to the Air Force Academy. But in any event,
this technology is being applied ubiquitously now. It's being used all over the place for
different things. We have to apply the technology in a way which is ethical. We have to make sure
we have human control over our lethal means, ultimately, so that we don't do anything that
violates the laws of war. But we're ready to move forward with some significant steps
at this point because of where the technology has matured to.
Talk about ethics. How do we do that on the global stage?
First of all, we have to design those requirements in. They're part of the design process.
And secondly, we have to move fast. Because of just what you said, people who are not constrained
by that, who don't take the time and the effort to design those kind of constraints onto their
systems, do have, in some ways, an advantage. I think overall, we have the high ground here,
and I think it's the right thing for us to do. So that's what we are going to do.
But it dictates that we move as quickly as possible to integrate these technologies
into our systems. Again, in a way consistent with our values.
A strong offense is the best defense, right?
We're in a race right now for military technological superiority. We've got to
move quickly. You don't win a race by standing still. The response I've had since I introduced
the Ungrouped Collaborative Combat Aircraft program over two years ago now has really been
fantastic. Industry's responded very well. Our government system has responded very well.
And we're moving forward very quickly.
Okay. The machine is making some decisions, right? But are we always going to have a human
involved in that kill chain? Are we ever going to get to a point where
the machine is actually making that lethal decision?
We will always have humans involved in the application of force. And there are questions
as to exactly when in the chain and how much of that decision-making process is delegated
essentially to the machine. But even if it is, it has to be delegated in a way which is consistent
with the rules that we have to follow. And we have to have a way to hold humans accountable
for the performance of that machine. So, you know, at the end of the day,
humans who put that machine on the battlefield, who supervised it, who designed it,
are accountable for its performance. And we've got to maintain that. I think that's central to the
rule of law, basically, and to ensuring compliance. Well, that means a number of things. One of them
is that we have to make sure we put the right rules in place and we have to test to ensure
that they're reliably enforced. And I think, frankly, as we get into this, we'll see machines
that are better than humans at both doing the operational tasks they've been asked to do
and at following the rules of engagement. So there are some things that the machines
are probably going to be a lot better at than humans. I mean, in the airplane,
they don't have to worry about how many Gs they pull. They don't feel the way a human feels,
right? So obviously they're going to be a lot better at some things. But what are those things
that you see humans always being better at than machines? For the foreseeable future,
humans, I think, will be better at making broader judgments, at assessing the totality of a
situation, and determining whether or not they're on a path to success or not, and altering the
strategy or the play. The concept that we have for the uncrewed collaborative combat aircraft
that we've started on is that a crewed aircraft, a piloted aircraft, will have control over more
than one wingman, if you will, which will be autonomous. And that pilot will call the plays.
Think of him as a quarterback, if you will. And he'll be able to call an audible, if he needs to.
He'll be able to adjust those plays, depending upon the tactical situation,
and how well he thinks that team is doing. That's the basic concept here.
But he will have to ensure that, given the rules of engagement and so on, we're not violating any
of those rules. We're not engaging aircraft, for example. We should not engage. And he'll
be supervising, if you will. I think that's the place we're going to be for some period of time.
In the distant future, we may get to even higher degrees of autonomy than that. But that's where
we're headed right now. So we're building loyal wingmen to fly on the wing of the human
fighter pilot. That's a good way to describe it. But I'd say it's a loyal formation, as opposed to
just a single wingman. Our work so far has shown that this is manageable, that the degree to which
the actual operation of the aircraft itself, that's become highly automated, too. So there's
less tasks that the pilot has to do to actually fly his own airplane. He can spend his limited
attention span and so on, making sure that that formation is operating as optimally as possible.
So we've gone from a place where we're able to test pretty regularly these different forms of
artificial intelligence in the air. Where are we as far as being able to start testing
the weapons with the AI? We're close to that now. We're virtually there.
The program we have, it consists of three parts. There's the uncrewed combat aircraft themselves,
which is the new aircranes that are going to be the loyal wingmen in that formation. There's a
program which is taking existing fighters, F-16s, and modifying them to operate autonomously,
but with a pilot on board in case the pilot needs to take over. And then there's a program to acquire
some existing uncrewed aircraft and put them into a unit and start to gain operational experience
with what does it mean to actually integrate those kinds of capabilities into a unit. So we're doing
all three of those things at once, so that we'll be fully prepared, mature the technology in
parallel with the development program, and be ready to adopt this new concept, if you will,
into our units as quickly as possible. What do you see as the greatest potential benefits
moving forward when we have AI fully integrated onto our battlefield? That's happening right now.
AI, again, is a very broad term, but we're integrating artificial intelligence in various
forms into our systems today for a lot of different applications. It's all being done
to be more effective and to be more cost effective. You know, the uncrewed combat
aircraft are going to cost a third or a quarter of what a crewed fighter would cost,
and they're going to give us better results on the battlefield. We've done a lot of simulation
to show that that's true, so we wouldn't be doing this if it wasn't going to give us an advantage.
That's the whole reason behind all of this. We can test a lot quicker now because we don't have to
be concerned with the same level of risk. Yeah, we can test through simulation, of course,
modeling, but we can also do live tests. There are a number of benefits that will come out of this,
and this is really the first major adoption of this type of technology and the first major
integration of it into our units for this purpose. We've had uncrewed aircraft that we've used for
ISR in particular for some time, and so we've gotten used to doing that. So that was one of
the steps down the path towards the kind of operational capability we're talking about now.
You mentioned some of the previous aircraft that we have that have been out there operationally
for a while delivering weapons and doing surveillance without humans on board. What's
so special about where we are right now? What's changed? What I observed when I was out of
government for about four years before I came into this position was that a number of things were
happening that had moved the technology forward to where we were ready to make a commitment to
this kind of capability. There's risk in that. When I told the Air Force over two years ago,
we were going down this path. We're going to feel the capability. We're not doing experiments now.
We're getting on with fielding the capability. It was based on work that had occurred over the last
several years before that. The ACE program at DARPA, Defensive Antiresearch Bystander Agency,
demonstrated that machines could actually outperform humans in dogfighting in some
situations. The Air Force's Skybird program built a number of prototypes of uncrewed aircraft that
could lead to where we're going to go. The Australians, working with one of our contractors,
Boeing, developed a loyal wingman program that they brought forward. So there was a good body
of evidence, as far as I could tell, that we were ready to make the commitment to an actual
operational capability at scale. And that's the path we're on now. What would be the danger of us
not pursuing AI at the rate at which we are now? We're in a race for military technological
superiority. And if we don't move forward, we're going to lose. And this is a substantial move
forward. It's not a mystery that this is a possibility. So we can expect our adversaries
to go down the same path at some point, if they're not already. Critics would say that there are
concerns with the potential of things going wrong with AI taking over, if you will.
We have to design very carefully, and we have to test very thoroughly to ensure that we don't
have something like that happen. We also have to make sure that the other side can somehow take
over control of these aircraft. So we've got to design all that in. I think we know how to do that.
We're aware of those risks. And one of the things that has surprised me a little bit,
quite frankly, is how our fighter pilots have embraced this. I thought there might be some
resistance to this idea. But they recognize that a modern battlefield is going to be an intensely
lethal place, no matter what. And they would much rather have some of these uncrewed aircraft
out in front, taking fire so that they don't have to. This is going to save a lot of pilots' lives
when we get this implemented. And our pilots are aware of that. Yeah, as a fighter pilot myself,
the thought of being able to fly out there on a mission with other aircraft where I don't have
to be concerned about a human in that plane is extraordinary. Jackie, that's 100% correct.
This will give us a range of tactical opportunities that we don't have today. It will allow us to put
some of these aircraft at risk. We wouldn't want to put a crewed aircraft at risk and risk a pilot.
Our experimentation, our simulation work shows that there's a very high payoff to going down this road.
So our kids, all three, want to become fighter pilots someday. No surprise, mom and dad are both
fighter pilots, retired fighter pilots now. I'm really curious to see what the landscape is going
to look like for them. What do you envision for them 10 years down the road or so?
What do you think that that battlefield is really going to look like? For the foreseeable future,
I foresee a mixed force. You want to have high bandwidth, low risk lines of communication,
ways to control these aircraft, and again, call audibles if necessary. You want a human on scene
with a situation awareness to do that. You continue to have crewed aircraft fighters, and you continue
to have a way for them to control these other assets. I think that's a model we're going to have.
I don't see them as necessarily reducing our force structure for fighters. I see them as complementing
that and adding to that, and I think that's the path that we're going to be going on. Sometime in
the far distant future, maybe somewhere else, but as far as I can see right now, that's where we're
going to be. I am raising little fighter pilots back home. I'm a wannabe fighter pilot, and I fully
understand where you're coming from. I'm going to get a ride shortly in one of our test aircraft
that's testing out this technology. I'm really looking forward to that. You've been a proponent
of autonomy and AI for quite some time, and it's easy to think about how it's used
on the battlefield as wingmen, as fighter pilots out there, but what are some of the other,
maybe less flashy ways that we might be considering artificial intelligence within
our military? Basically, what we have is significant advances in computing power,
and data storage, and software that operates, so we're able to fuse a lot more data a lot more
quickly and make a wider range of decision-making, if you will, support from the computers, if you
I know we use the term AI an awful lot these days. I'd much rather talk about the specifics of what
it is we're trying to do, because again, it's not really intelligence per se. It's the ability
to do certain functions that we associate with human situation awareness and inference,
and to do it in a way which is much faster and much more accurate. That's the advantage that
comes from AI. Part of it is the operation of the actual aircraft, but much more significant
is the ability to bring in data from multiple sources, integrate it together, and draw inferences
from that, tactical inferences in this case, and then support tactical decisions, including
engagement decisions ultimately, but all understandably for the reasons we've talked about
under human control, so that you don't have anything go wrong, and we're going to wring the
heck out of this stuff. We're going to make very, very sure that it works in a wide variety of
circumstances, and we're going to make sure that that control is maintained as operationally,
and then we'll learn. We'll go for it. There are a number of other areas where what's called
artificial intelligence has applications, how we do planning of operations, as well as conduct an
actual air fight, battle, if you will, how we structure maintenance and do things associated
with maintenance and training, how we get the most out of that. In the case of the human-machine
interaction, there's the question of, is the pilot interacting as efficiently as he or she
could be or should be with the autonomy that's available to that person, and then giving feedback
so that that person becomes better and better at that. One of the things that what's called AI does
is called machine learning, where you basically do millions or trillions even of operations with the
software to get better and better and better at trying to get the outcomes that you designed it
to achieve for you. Integrating humans into that is going to get us overall much better performance
from both the machine and the person who is involved. The person, when you think about it,
the instructor, no matter how much experience they have, is limited in comparison to these
machines that are able to very quickly draw on all of the data that's out there, all of the
potential scenarios, the potential outcomes, and help us learn at a much quicker rate.
Jackie, you know this better than I do. Time matters. Seconds in fractions of seconds matter.
I've had a lot of interactions with pilots and people who are involved in air combat. A human
being at his best is going to be a few hundred milliseconds making a decision about almost
anything. Machines will do it in a very small fraction of that amount of time, and when you
add that up, you get a significant advantage when the human takes on the role of overseeing and
supervising the machine, and the machine does all those calculations at blinding speed compared to
what a human being can do. You think about not only how can it help fighter pilots become better
fighter pilots quicker, but the maintainers on the ground looking at the systems and the airplane,
knowing what's going to break before it breaks, and then to think about things like
the bigger picture, the strategy, the tactics, the war gaming.
Exactly. You can run analysis for a tactical situation of lots of courses of action. Even
more importantly, perhaps, you can also adjust as things in reality change. You know, there's, of
course, a well-known saying that all plans, you know, are obsolete at the time you, you know,
you reach the line of contact, right? And then you're dealing with real life instead of what
you imagined ahead of time. This technology will allow us to make real-time adjustments to changes
in the actual tactical situation much more quickly and much more effectively, I think,
than human beings can do that today. The United States has made rapid advances
in this kind of technology, and we're taking our time to do it right, to do it ethically,
but what happens if an adversary starts deploying this kind of technology sooner? What do we do?
We're going to be at a disadvantage, and I think we'll have to rush at that time to try to get
into production and fielding quicker. I think it's important that we limit our reach for our
first increment in particular, that we get a minimum viable product, if you will. We get
the set of capabilities that are most important to us, get those fielded, and we're going to learn
a lot from that. The way the program is structured today for the uncrewed collaborative combat
aircraft is in two increments right now. The first increment does what I said. It gets substantial
numbers fielded within the next few years, but then right behind that is the next increment
that will have more advanced technologies, additional functionality, and another level
of capability compared to the first one, but we're going to learn an awful lot. We're going to give
ourselves a real capability with that first increment that's going to be operationally
important to us. Do you feel like you have the support you need? Absolutely, I do. I have good
support from the Secretary of Defense. I have good support from the Chiefs of Staff for the Air Force
and from the entire team, including the operational community that you were part of.
Yeah, for me as a fighter pilot, it's very exciting to think about the possibility of being out there.
Is massive force kind of one of the first places that you would go, or are there other avenues that
you think are a higher priority? Getting meaningful military capability fielded is what I care about
the most, and I care about our pacing challenge, China, the most as an adversary. You know, Russia
just demonstrated the limitations of its own military pretty directly, particularly its Air
Force. China has been working very hard, and this goes back at least 20 years, to field the military
that can go head-to-head with the United States, and first, if they can, prevent us from intervening
in the Western Pacific, and then defeating us if we do. They've fielded a number of systems that
are designed to take on particularly our space and our Air Force or airborne assets, wherever they
may come from, and we've got to respond to that. They're moving forward aggressively. The disparity
in time between when we can feel this kind of capability and when they have something comparable
is not going to be very much, whichever way it comes out. AI has become such a popular topic as
of late. Most people think about ChadGBT and using it to help write a better story or make social
media posts and things like that, but it's been around in development behind the curtain for
quite some time. Where do you see it going soon, and where do you see it going over the next, you know,
10 to 20 years? It is a technology that's moving very rapidly, and you mentioned ChadGBT. It hasn't
been out very long, and already, you know, refinements of that are occurring at a really,
you know, mind-bending rate, and I see the same thing in other types of AI. One of the key,
one of the CEOs, one of the major software companies said that AI was just another term
for software that works. I think there's something to that, actually, and again, I think we need to
talk about the specific applications and whether the technology is matured enough to take on those
applications, and that's why we're moving forward very quickly on collaborative combat aircraft,
because we think that that set of applications is ready, and we can go ahead and move forward.
There are going to be others. I don't think warfare will ever be entirely automated.
It's a very messy, ugly, complicated, confusing endeavor, and human beings are always going to
have to have a role in overseeing it for a whole variety of reasons, but I think automation,
which is a term I would prefer, will continue to be part of the equation, just because of the
inherent limitations on human beings. We can only handle so much in our minds, and we can only process
it so fast, and the machines are going to do better, but they've got to be accurate, and they also have
got to be consistent with our values, and our ethics, and our legal obligations.
There's inherent risk pursuing autonomy and artificial intelligence in the world of military
defense, but there's a lot of risk if we don't pursue it.
That's absolutely correct. I don't want us to wake up one day and discover that we're at a substantial
military disadvantage because of what our potential opponents have done. We need to prevent that.
There's no question in my mind that this is the direction in which technology is taking us,
and there's no question in my mind that it's going to confer a substantial operational advantage.
We've been talking about the error logic. There's also the question of space,
which has been militarized by our opponents to a large extent. It's been operationalized,
and one of the things that we have to be concerned about, because of the vulnerability of our space
systems, is the other side acting in a way which surprises us, and which we can't react to on the
timescales necessary. We're going to need automated capability to respond to that kind of an attack.
It's going to happen at a pace that, again, similar to a fighter pilot engagement, humans will be at a
huge disadvantage relative to the machines. There isn't any question in my mind that we have to move
forward with this, and move forward as quickly as we reasonably can. It's a game changer,
and we cannot afford to be second in the race.
Secretary, thank you very much for your time today. I really appreciate it.
Thank you. Great to be with you.
We hope you enjoyed this conversation with Secretary Frank Kendall. To hear more full-length
aviation and space interviews, check out the Behind the Wings podcast wherever you listen.
And to see this technology in action, check out our newest season of Behind the Wings on PBS.
All the links are listed in the episode description. We couldn't cover everything,
so leave your questions under the video, and we'll answer as best we can.
If you subscribe, thank you so much. If you don't, just subscribe already.
We'll see you next time, right here on Behind the Wings.
