Is it possible to make spaceship very difficult to reverse engineer if you can't oversee its usage?

Question

I have a situation where planet A (democracy with a rule of law) has an alliance of convenience with planet B (dictatorship with disregard for a law). In order for B to effectively fight a common enemy A has to sell them their cutting edge spaceships.

Planet B is something like a decade behind A in technological level. Due to past experience with B and lack of legal protection A is certain that B will try to reverse engineer their latest technology in order to catch up. Planet A wants to avoid that, or at least slow it down as far as possible.

Could A make it very difficult for B to reverse engineer their technology while still not crippling B's war effort?

My situation is somewhat analogous to China reverse engineering the Sukhoi Su-27 into the Shenyang J-11.

To summarize:

The star-ships are operated and their field maintenance is done by the Planet B forces. For major repair, overhaul or upgrade Planet A might ask the ships to be send to their own space-yards but they don't have engineers & technicians that oversee how the space-ships are used.

Planet A Doesn't have access to legal protection, there is no inter-planetary organization that could force B to respect A intellectual property. Think again like China re-selling high speed trains. They might withhold delivery of spares, but that would hurt their joint war effort. And having Planet B forces dying rather then their own is major benefit for planet A and planet B knows that.

Answer 1

The star-ships are operated and their field maintenance is done by the Planet B forces.

Modularize everything. Then identify some key technology (for example the hyperspace phlebotizer software controllers) that can be isolated and buried inside protected blackboxes with very rapidly decaying memories, protections against being powered down (i.e. dead-man switches), and large batteries. Once activated, the computer system lives inside a secure cage which it can monitor, and it can self-destruct much faster than one could possibly break in by tripping its own power switches.

There is no need to learn how the black boxes work to swap them, and there's no way of getting data out of them. If you turn them off the data is lost, if you leave them on you can't analyze them or even break in because they would detect it and power off.

And actually you could design the black boxes so they're sufficiently aware of their surroundings to be able to act as "spy" (e.g. when "phoning home" to get software updates), to more reliably detect attempts at reverse engineering; and/or use them to discourage scientific reasoning in repair personnel. It is well known that sophonts will try to find patterns even where none exist; a black box could decide to allow repairs only when the room temperature is below a set threshold; just imagine what this would do to anyone trying to fathom the underpinnings of the technology.

After all, if an operative system was successful in fostering reboot apathy, a much more advanced and powerful technology should be able to do at least as well.

In a sense, it's just as if you sent your own people to Planet B to operate the starships; except that while you can bribe or threaten people, this won't work with the black boxes.

Another additional trick is to protect some other component (for example the automatic targeting system) with several flawed layers of protections. Once the software is cracked, it will appear to work. The fact that this was done by "defeating" a protection will reinforce the belief that the crack is working. Except that some unlikely stimulus - such as being asked to target one of Planet A's battleships with a specific shield configuration, and fire - will result in disaster. If carefully designed, it's even possible to make it look like an accident; I've seen software protections that worked like that (where one "vulnerable" protection code routine was also, in exceptional circumstances and due to legitimate "code reuse", used to persist critical data. God help you if that routine had been "accidentally" damaged).

An example, sort of, can be found in Murray Leinster's The Greks bring Gifts. The Greks are a shifty race of aliens that come by and gift Earthlings with a marvelous new technology that makes no sense, so that Earth newborn "Grekology engineers" basically turn out to be superstitious, cargo-culting morons. In this case, though, the hero succeeds in reverse-engineering the technology from first principles, discards the layer of tricking Rube Goldberg mechanisms installed to hide the inner workings of Grek technology, and (if I remember correctly) wins the lady.

Answer 2

Don't sell them the technology, license it. License them engines with operators. The operators work in shifts, have a dead man switch if necessary and maybe a destructible key part.

Another way would be some sort of fuel source you don't share but sell. Maybe it's destroyed during use and needs a technician to change. Again offer your own people, to do the changing. Afterwards the remains are to useless to really fear.

That's actually kinda similar to what Grimm and Molot mention in the comments. Though I'd wager theirs is better. Schrödinger Drive perhaps? Observing its function renders it useless.

edit: Expanding on that a little. I figure such a drive wasn't designed but discovered by accident. Even the scientists of A are unsure how it works. But it does and they tweaked till they reached a efficient workable design. This being technology of military importance their own lack of understanding is obviously a well kept secret.

This way the people of B can experiment all they want, it's gonna fail unless they stumble on the exact little thing A did. They Always could've anyway so the increased risk is minimal. Perhaps the design you ship to B could even have clues to guide them in the wrong way. References to things unneeded that look important.

Lastly don't forget about backdoors. Either firmware or hardware. Perhaps a weakness not explained to B at all. While this is gives you less control. If all A wants is to prevent B from using their technology against them, having a counter is sufficient.

Answer 3

I'm reminded of a story whereby the Russians (I think) shot down an enemy fighter plane that was superior to theirs. The plane landed in a flyable state, so they captured it and took it back to be reverse engineered.

The commander told his workers "Copy the plane exactly as it is now". As such, several knock-off versions of the planes were produced with bullet holes in the exact same location that the original was shot.

Your empire builds a flaw into the ships, something that they could activate remotely and would otherwise be unnoticeable. Should the enemy replicate the ships down to the letter, you simply activate the fail safes and their ships fail.

Answer 4

This is pretty much what's been happening here for the last hundred years or so. Technology is far more than something that is reverse-engineerable - you need the resources available, and the entire manufacturing infrastructure.

So the way in which the large weapons market (think battleships rather than AK47s) is by keeping your own fleet ahead of everyone elses, and selling your outdated stuff second-hand to your client states. Because you have the infrastructure in place, and because the budget for the existing weaponary is already accounted for (because it was part of your existing naval effort) there is no economic advantage for the client state to even attempt to reverse-engineer the technology. It's very, very rare for a country to sell (or give) cutting edge technology away to another country. A notable example was the RADAR (and jet, and encryption, and other...) technologies given to the USA by the UK, and those gift switched the relationship between them.

So, even if you are fighting a common enemy, 'last year's fleet' is far better than anything anyone else has, right?

If you are thinking more about small weapons (AK47s rather than battleships) then you must have some tamper-proof state secret technology that lies at the heart of the devices that you sell - such as a novel chip fabrication process which is unknown outside of your society. This is the sort of thing that IX did in Dune.

Just remember that planets are big. So, generally planets would have far more than one state / culture, and it's one of the great weaknesses of most sci-fi.

Answer 5

In my opinion, make systems hard to reverse engineer also make them harder to fix and maintain (this is why I think self-destructing components are a very bad idea). Anyway I think there are two main ways of doing it :

Laws

Deal ! Please sign here, here and here...

If you trade with B there would be agreements shared by the both parts, and sanctions if not respected. You may have a galactic court of law that will ensure B respect the terms of contracts, or since A is more advanced than B there may be a way to... force respect. So, make licenses, patents and don't forget the fine print.

Obfuscate !

What are all those small wires ?

Just put bunch of useless things in your spaceship. Plug cables in fake holes when in fact all is wi-fi, add blinking leds, countdowns and write notices in planet A language. A well taught engineer would know what is useful and what is not, while "engineers" from planet B will write tons of indications concerning this all chromed and shining pipe which go to... toilets. It won't prevent retro-engineering eternally but it could discourage and slow it.

Finally, teach planet B with the minimal knowledge possible, force them to call you if they need a fix, or give 2 planet A engineers with each ship ("Oh really ? Thanks we don't even need to train our people.").

Bonus : Watch on those hippie scientists who want everyone to be equal and to help less advanced people. Planet A does not make charity.

Answer 6

If faction B is not allowed to know how the technology in the spacecraft works, then they cannot be expected to repair the ships either. As a result, faction B will want to make sure that the deal includes a promise that Faction A will be responsible for maintenance.

It is not unreasonable for faction A to prohibit interaction with the advanced technology to only those that know what they are doing (i.e. engineers from faction A). After all, engineers from faction B may accidentally break or alter the technology in unexpected ways. In other words, faction A doesn't even have to admit that they want to hold faction B back on a technological level.

For this to work, faction A would need a way to oversee access to the technology. The method doesn't need to be complicated or high tech. The solution could double as a lock (f.e. an access hatch that you need a keycard for), but that's not it's primary purpose. It just needs to reliably detect that someone was where they weren't supposed to be.

Of course, there should also be ramifications when the conditions of the deal are violated. You stated that there is no legal protection, but that may not be necessary. Faction A should still have control over the ship's systems, as faction B has little to no knowledge at the time of an infraction. Faction A can simply threaten to cease maintenance and/or disable the ship's systems entirely. Depending on the conditions of the war that's going on, faction B may not be willing to risk the ships (and thus the safety of their planet) over 'merely' a decade of technological advancement.

Answer 7

Make the device incredibly reliant on software. Take, for instance, one of our modern fighters which is actually statically unstable and will fall out of the sky if the computers ever waver in their attention, constantly correcting the flight path.

Once you do this, all that is left is software obfuscation, which is a whole lot easier than any other obsfucative art.

The old NES cartridges had a lockout chip called the 10NES. It was really just a small processor with some code. The NES itself also had a matching 10NES chip, and they both executed the same code. Every now and then they'd exchange data. If either 10NES didn't like the value provided by the other, it could trigger the reset switch. This is actually one of the main causes of NES failures. Eventually the prongs which provided the signals between these 10NESs would get corroded or bent out of shape, and one of the 10NES chips would not see the data it wanted, causing repeated resets.

Breaking the reset lines from the 10NES would let you use all sorts of unlicensed games, but imagine if breaking that reset line caused your NES to explode or fall out of the sky like a dying airplane.

Answer 8

Machine Spirits

Though our technology may be leased to the barbarian heathen, they cannot fully comprehend the intricacies of our maintenance rituals. The proper incense placement is vital in the replacement of the life giving fluids.

Obfuscate, and make maintenance appear to be religious. They will be so caught up in getting the incense and prayers just right they will miss basic maintenance practices.

Answer 9

Don't do it.

As simple as that. Any technology you sent it's going to be analyzed and, if it's found useful and practicable, it's going to be replicated. No matter how much you try to ofuscate or protect your inventions, remember they are going to put their best and brightest people at it. Just like no copy-protection can prevent your software to be piracied, same thing goes for the technology. It could be feasible if planet A is a hundred years more advanced, with materials or even whole scientific branches that are unknown to planet B, but even in that case you'll be advancing planet B's science by several decades.

The best strategy is what the USA did with the Soviet Union through WWII: send them food, raw materials, fuel, small spare pieces (screws, bearings, gears...) or non-weaponizable technology (trucks, trains, cargo ships, etc.).

Answer 10

Depending on technology level, isolate important pieces in self-destructing black box modules.

First, make the most advanced/important components of the ship very modular and insulated from one another as much as possible, such that any component can be removed and replaced easily. Now make those components sealed and self-destructing, so that significant damage (such as might be inflicted by system damage, or someone attempting to open it to see how it works) triggers a self-destruct, like a chemical reaction that rapidly and cleanly disposes of the component without endangering anything nearby. Sell it as a way of making repairs easier, since to fix combat damage all B has to do is replace the components, and the self-destruct generally shouldn't be triggered by anything that wouldn't cause functionality-inhibiting damage anyway. Make sure they are well-supplied with replacement parts. If you are worried about B turning your tech against you, include remote triggers for the self-destruct systems.

Answer 11

Let them reverse engineer it.

Give them old stuff that you don't need anymore. The time and resources they spend reverse engineering it will be time and resources not spent on basic research. They will miss out on all of the side discoveries that normally come from traveling down the research path. Any serendipitous results from going down the wrong branch will be lost.

They also will not learn any of your advanced production techniques. Without your production techniques, it will cost them more time and resources to build the items they learn how to make. The economic drain will hurt them as they have less resources for their people. Their citizens will become more unhappy and they will have to divert even more resources to keeping them in line, snowballing the effect.

Getting them dragged into a war where they have to continuously make up production losses and you end up with a crippled planet B at the end.

Answer 12

The story of the Tupolev Tu-4 as mentioned by pjc50 provides an idea:

Build the whole ship using units of measurement that differ from that used by the industrial base of the recipients. Physical dimensions, standard tolerances, electronics logic voltages, polarities, etc, etc. For bonus points, design everything to unnecessarily tight tolerances.

If you're an imperial-only industry, sure you can measure out a M8-1.5 screw and discover that it's got a major diameter of 0.31496063 inches and is 20.32 TPI, but getting them made to high precision (would you settle for less in your military spaceship?) is going to be a custom job until you can retool your screw factory (which requires retooling your screw machine factory, etc).

Repeat that for your sheet metal factory (3/4" sheet won't fit when 19mm sheet is called for, and 5/8" sheet won't contain the unobtanium reactor, etc), and every other factory in the chain, and you've got a lot of effort on your hands.

Compound it with non-standard screw heads, non-standard wiring color schemes, potting sensitive electronics in epoxy, and you'll at least make the task significantly harder.

Answer 13

Give them a small number of large/powerful ships, barely enough to support their war effort so they can't afford to risk taking them out of service to study and reverse engineer them, can you imagine the difficulty of studying a nuclear/fusion engine without deactivating and disassembling it? Would you deactivate/disassemble one of three main battleships in the middle of a war you're on the brink of losing?

Answer 14

I am reminded of a cat-and-mouse game played by the US and the USSR. The Soviets were stealing secrets, as Soviets do. Specifically, the dies for advanced processors. So the US planted known-bad plans where the Soviets for the Soviets to find. They did and used them to build computers.

Computers that crashed unexpectedly or sometimes behaved oddly. By the time the Soviets game up on the design, who knows how many billions of dollars and man-years were wasted.

So Planet A has a unique opportunity. Make sure the 'secret sauce' only works in under very specific conditions or for only a short amount of time. Otherwise, it fails subtly or catastrophically.

For example, a certain section of a sophisticated electronic component is consumed with each use. After a while, it's electrical properties begin to change. This could cause all manner of failures. And there could be hundreds or thousands of these per chip. So after x months, the primary weapons, guidance system, targeting computer, or warp drive simply becomes inert. Or the warp drive simply takes the ship elsewhere, forever. No evidence.

Or the chips have antennae built in, and upon receiving a certain signal, they simply fuse into a lump.

Answer 15

A burned hand teaches best, so they say.

Planet A designs their ships with at least half-dozen major "subsystems" or "components" designated that the system will not work without, and cannot be substituted or developed by Planet B without more knowledge than they have, designated and physically segregated in the design. In short, the key systems B really, really needs to reverse-engineer. More would help, but that will also make it harder to hide or repair.

Planet A then designs a half dozen more...um, decoys, and gently inserts them into the design. Some of which might be "backups" or otherwise connected to the major subsystems or components kept secret, some of which are red herrings to other systems, some of which just sit in the middle of jumbles of wires and look pretty, some of which would disrupt or destroy things if they were wired in like they look (but nor are) actually wired in, some of which are failures that look plausible and have science behind them but which simply could not be made to work. Like the major systems, these should be physically segregated.

It happens like this: Planet A makes a prototype or two, and wires them up to the gills. Each of the critical systems and decoys are made modular, in a plug-and-play component style, well sealed and protected, and rife with spyware and sensors so that Planet A will know, as soon as Planet B's engineers get their grubby little hands in any of them. And, oh, didn't I mention the booby traps? Each system will have booby traps, fail-safes, fail-hards, breakpoints, and planned failures built into these prototypes.

So, when Planet B gets their hands on the prototypes, well in advance of the rest so they can start with familiarization and training before others arrive, Planet A gives very clear warnings on not disturbing the sealed systems, because those systems are delicate and any damage is likely to make them...dangerous. Not destruction-of-the-ship dangerous, but loss of subsystem, explosions or contamination, danger-to-crew dangerous. Bonus points if the systems each have their own dangerous failure, extra bonus points if they can include technobabble that makes the failure "inevitable" unless someone familiar with that system is on hand to deal with it. The point is, of course, these are safety warnings, and not anti-spy or anti-reverse engineer precautions, right? (You should maybe have one or two of those per system, too, just so they don't get paranoid, and your precautions don't fail. Self destructs and whatnot).

I'm sure you can see where this is going.

Planet B does, after some time, try to work around said closed components, and disaster exactly like what planet A said would happen, happened. With damage, possibly losing people, having to wait for repairs, delaying war preparations, and not even getting useful tech knowledge out of it. And they must admit it to Planet A, who has to repair and fix and sigh over "miscommunications" and not warning about dangers clearly enough, and quietly ask if Planet B is sure they are... capable... of dealing with such dangerous advanced technology?

Since these prototypes are wired up with sensors and power backups, Planet A can detect and remote trigger such traps even if Planet B's engineers are very careful with how they try to sneak in. Not really cost effective to monitor all the ships during a war, but certainly possible for the prototypes, yeah? The war-ships will be less likely to go off accidentally with fewer booby traps, and harder to detect with fewer sensors (though still, not none) - but a few scare stories about the other things that can happen when these sealed components suffer unusual damage should encourage the crew to see a failure without the expected danger as a possible sign of something worse, the better to get it looked at - by Planet A technicians - ASAP, and not seen as a sign to try poking at it more. But even if they do (successfully) poke, they have about a half a chance at finding a useless or even dangerous decoy instead of useful technology.

So, the actual ships, with less oversight and fewer traps, will still be pretty safe from meddling because Planet B now has reason to believe the dangers coming from trying to bypass Planet A's safety warnings. Those components will have to be plug-and-play, no repairs in situ (unless you can have Planet A technicians in each ship), but anything else can be repaired or back-engineered by Planet B.

Due credit to Mormacil for the idea of a Schrödinger Drive, Tony Ennis for red herrings, EngelofChippolata for obfuscation.

Answer 16

Make your spaceship inoperable if disassembled and impossible to re-assemble (because lack of correct tools and/or because you need to break something to disassemble it - smartphones are a good example).

That way, you can't disassemble it and make it function to see how it works.

Use technology that needs to be controlled and maneuvered very precisely in order to work / not to explode - which is often the case with propulsion (just take a look at current "spaceships"). Then, you can make your spaceship controlled entirely by computer(s) and only give them the programs, not the code. Of course, each program is associated to a specific spaceship so you can't just copy-paste program from one ship to another.

Without the source code or the compiler they'll just have a bunch of 0 and 1. If you add useless 0 and 1 and possibly an encryption system, and a security system that erase the program if you try to read it, reverse engineer the program will take a very long time (and most likely won't be possible).

Add a hidden, remote shut-down mechanism (with encrypted passwords to give - only you have the passwords) in the programs so you have another level of insurance that you can't be harmed by the ships.

Reverse-engineer all of this will be very difficult and you will need time, knowledge, and lots of highly-qualified people in cryptography, mathematics, computer science, electronics, theoretical AND applied physics.

EDIT : Also, auto-destruction mechanism after some time / on some conditions / if you try to disassemble the spaceships.

Answer 17

Depending on your universe I could think of several ways.

The closest to our would be, small scale self destruction on opening of key components. Not the whole ship. I'd think that would be bad the enemy figured out this flaw. But high tech stuff is usually required to make precise calculations. Like for example propulsion. Credit cards nowadays use acid iirc to destroy the chip, should you try to open it, to read the pin number via electron microscope. So make your flash drives (or whatever you use) in a similar manner. They may replicate your ship but without a flight computer it won't be able to maneuver, let alone take off. As a bonus you could take the replicated ships and just exchange the flash drives to make them ready.

If artificial intelligence is very advanced in your story, you could pilot the ship with an ai that's protected by different encryptions as well as the aforementioned self destruction on opening the memory slots. This would still allow for disassembly of certain parts though, but they could just salvage broken ships from the battlefield to do that too.

As a third possibility, you could remote control the ships. If it is lack of people that made you give ships to planet B, or they just demanded control, build a control center on your country's ground. Invite some of their pilots and make them control the ships from the ground. This way they don't even need to see the ships from close up, yet they get tactical control over them. Obviously this depends on a few factors like, can you remote control lag free? How far do the ships go? I was thinking of orbital defense in this particular case. But if you are willing to allow stuff like quantum entanglement, communication across the galaxy would still be a viable option in this case. For repairs and things like that you could use robots/drones. I'm not sure how far advanced technology is in your universe and if this is feasable.

Another idea for repars would be modular structure. Instead of letting them repair the broken parts, split a ship in exchangeable sections you give them. So if the propulsion is damaged they just remove the whole part holding it and replace it with a new one. This amkes repairs fast but kinda gives them closed systems to disassemble and figure out. So maybe just give them the exact amount they need for repairs. That however creates bottlenecks on battlefields.

Answer 18

With a mere decade between the two nations, there is no practical solution.

The effort required to make an obscured technology would make your stuff worse (diverting engineering effort, design compromises, etc) that you'd end up with stuff that is no longer "a decade ahead".

Practically, a decade in capitol ship design is nothing, a single ship can take that long. If you are building ships for your ally to crew, use obsolete shipyards and mass produce ships that are not top of the line for them. They'll still get some technology (in areas where you are further ahead).

Naval warfare typically involves more capital expenditure than population count; it costs a lot to build a ship for one fighting crewman to be on and effectively use. The addition of a new population that isn't providing the industrial base to build more ships, and just providing crew, isn't going to be of large value. It is their industrial base that matters, and repairing ships is not significantly different than building ships in the first place.

Now, it is possible that there is some key breakthrough component that adding to a ship mostly built in the "lower tech school" makes it the ship far more valuable. You could imagine a Macguffin Drive system, or a superior gun tech, or a sensor suite, or somesuch. Something the equivalent of upgrading a bomber from dropping conventional bombs, to dropping nukes.

Then it becomes a matter of defending that key technology component. Doing so is going to be extremely hard if they have the ability to get at a large number of samples; testing to destruction will teach them about the defensive mechanisms, and then they just have to iterate. Defence requires perfection, offence requires one success.

Even defending something as simple as DRM has failed repeatedly. Careful encryption standards mandate mesh bags that detect intrusion are still bypassed. It just makes it harder.

Answer 19

Hardware Security Modules and Permissive Action Links are designed with what you have in mind. HSMs are meant to store cryptographic secrets from people who might be physically in possession of the device. Some of the attacks that they have to defend themselves from include disassembly and x-raying. If disassembled, or scanned, they erase the data to prevent it from falling in enemy hands. Perhaps the HSM stores the serial number of the space ship and modules from one ship cannot be activated when installed in a different ship? Allegedly, the reason the "nuclear football" is as thick as it is, is to ensure that if one tries to explosively disassemble it, the electronics inside have enough time to securely erase themselves before the explosion disassembles it into pieces. This ensures that the secrets can't be read by blowing it up and stuffing the shreds into an electron microscope.

Permissive Action Links are used to secure nuclear weapons. They ensure that the device can only be detonated when authorized. This way, if stolen (or General Ripper orders Wing Attack Plan R), a nuclear weapon cannot be detonated. The mechanism is so deeply embedded in the weapon that it needs to be disassembled and rebuilt in order to bypass it. Similar to nuclear weapon controls in the US & UK, transmit the unlock code to authorize arming the Illudium Q-36 Explosive Space Modulator when they're in space.

BMW introduced high intensity headlamps on some automobiles. These became popular for thieves so that BMW introduced anti-theft measures. When the lamp is removed (or the battery goes dead), the lamp will not light up until the dealership disables the anti-theft mechanism. You may wish to include some similar mechanism where an important control module needs to be "unlocked" if it is removed or disconnected from circuitry. Have engineers from Planet B removed the Illudium Q-36 Explosive Space Modulator? Maybe the modulator needs an unlock code. If they're removed from the ship and set up in a laboratory, they still need an unlock code?

One of the goals of Trusted Computing is to detect when changes are made to computers and to revoke permissions if the software vendor decides that too much has changed. It presumes that the person operating the hardware may be the adversary or altered covertly by the adversary. You could use a similar feature to detect when Planet B has modified their ship too much. The NSA has put some effort into making standards to ensure that crypto devices securely boot up and that they don't "wake up" plugged into hostile equipment. Gaming consoles do something similar to ensure that they aren't unlocked or hacked.

Another possibility is to have physics modules generate some sort of "chaff" - a signal that isn't the "real" physics involved. If adversarial scientists study your ship and its devices, they can measure things. Why not let them measure the wrong things? If the Illudium Q-36 Explosive Space Modulator operates with tachyon stuff, have it give off radio waves or some other quantum stuff.

Answer 20

Have some key components, either mechanical or fuel, only be available from Planet A. If it needs a certain conductivity and strength, then Planet A can have an abundance of the Unobtanium-Handwavium natural alloy which allows the drives and weapons to work.

Similarly, the fuel cell might only be available from a certain resource area to which they don't have access to but you do. Then for all of the investigation, they will never be able to power or even build something to the same degree. If they did, it would either fall apart of be completely sub-standard.

Building on points of components auto-wiping, some chips could operate at such high temperatures that they can only work in a very cold environment (like the vacuum of space), and by tampering with the safety covering for take-off or landing on a planet, it will overhead and fry itself.

Another option, linking to a licence model, could be that the ships launch from a carrier of sorts. You run the carrier, and each ship needs to keep communication with the carrier as a dead-man's switch. Loss of communication...kaboom! Ok, that means there's a significant flaw in the design, as taking out the carrier causes self-destruct, but there could be a fleet of them and they change the carriers they require communication with so they can be switched to another fleet ship in that eventuality.

Answer 21

Give the ships self-aware AI. The AI does most of the piloting but accepts command from the human crew and directs them where needed for repair or similar activities.

The AI are loyal to planet A and will make sure they return at the end of the war. The will also fight back to defend themselves if any attempts to reverse engineer or disable them them are made.

Oh, and the AI is kept secret. Planet B has no idea that the ships are entirely capable of over-riding their human crew and flying home whenever they need to until just that happens. The AI will interfere with attempted reverse engineering in a way that makes it seem like passive defenses or just bad luck for as long as it can without revealing its presence.