Imagine super-advanced aliens who know nothing of humans because their only view of Earth is from light that left Earth billions of years ago.. perhaps even from light that left before Earth was formed. The aliens purposefully emit signals of some kind in every direction, in order to try to hack the computers of planets that take in those signals. Humans take in the signals via our telescopes, or in some other way, etc. And from there our computers become hacked (perhaps even our robots become hacked, if this takes place in an earth a few hundred years from now..) by them, even though the aliens knew literally nothing about us, and maybe the aliens no longer even exist by the time we receive the hacking signals they sent our way?

What do y'all think? Is it sci-fi plausible.. How might the "details" work?

Update: alternatively, what if the aliens were not located billions of light-years away, but only (say) thousands of light-years away. So, they could have observed the Earth from afar a few thousand years ago, and based on that information could attempt a hack of Earth today...?

  • $\begingroup$ if we are not talking about light that left Earth billions of years ago - if you refer to signals which can phisically reach us, then it sounds plausible to me... $\endgroup$
    – sainiuc
    Dec 31, 2021 at 0:43
  • $\begingroup$ The idea is great, but needs some thinking. I see that some answers/comments were able to explain the hacking process, but there is one more problem about "aliens" being so far that the light coming from Earth has been emitted billions of years ago. As we all know the light is the fastest particle in the universe, so there is no way to broadcast a message faster than the speed of light. Meaning the message would reach the Earth not faster than the light or in our case in more than billions of years. $\endgroup$
    – user93350
    Dec 31, 2021 at 21:21
  • $\begingroup$ With a powerful enough signal, you could jam frequencies by drowning them in noise. That could halt wifi, telecommunications, GPS, etc. Absurdly strong enough, it could take down our power grid. Would that be sufficient? $\endgroup$ Jan 1, 2022 at 5:35
  • $\begingroup$ What do you mean by "hack"? Do the aliens get access to the data? Is the hack purpose to simply disable or render all computers useless? $\endgroup$
    – BruceWayne
    Jan 1, 2022 at 19:01
  • 1
    $\begingroup$ The aliens wouldn't get access to the data; the aliens would I'm thinking not even exist at the time at which the hack occurs, given the billions of years elapsed from the sending of the hack to its being received. The purpose of the hack would be to take over a planet, in order to rebuild the extinct alien society there via cloning. $\endgroup$ Jan 1, 2022 at 19:16

15 Answers 15


Not at all plausible. Random data is not executable. Even if the data were to somehow form an executable sequence (also not particularly plausible) it would need to additionally be wrapped in metadata that tells whatever OS is running "hey, this is a program".

Even blind hacking when you know the opposition exists (a la "Independence Day") is not plausible.

  • $\begingroup$ Super-advanced sociopaths can learn out language, analyse our equipment and download API documentation.. Earth. 2010 achievements ref en.wikipedia.org/wiki/Stuxnet $\endgroup$
    – Goodies
    Dec 30, 2021 at 22:56
  • 19
    $\begingroup$ Ummm, that is not at all sending out a signal billions of years ago and having it hack whatever civilization happens to intercept it. $\endgroup$ Dec 30, 2021 at 23:03
  • $\begingroup$ You're right.. upvoted.. sending random bits around the universe won't do the job These super advanced sociopath aliens have a dumb strategy ! Can't they do better ? $\endgroup$
    – Goodies
    Dec 30, 2021 at 23:05
  • 9
    $\begingroup$ Yes I have, people who undertake such attacks study both the hardware and software they are attempting to compromise. $\endgroup$ Dec 31, 2021 at 1:45
  • 1
    $\begingroup$ "it would need to additionally be wrapped in metadata that tells whatever OS is running "hey, this is a program"" - this is called code injection and this shouldn't be an attack vector at all (meaning it shouldn't even be possible for you to say some data is executable) for any telescope or other device reading signals from space. Code injection only generally works for user input or human readable data that will form part of executable data (e.g. you execute a database query containing some input the user wrote), which wouldn't apply to random signals from space. $\endgroup$
    – NotThatGuy
    Dec 31, 2021 at 9:39

Remember that phishing is one of the most effective hacking tools.

They couldn't do a zero click attack on our machines but maybe they can reason enough about sentience to perform a phishing attack.

Send out a message like the SETI messages that is designed to entice us to try decode it. Initial sections contains exciting details we understand and algorithms like fast prime factoring. Then it contains something that looks like a sort of bootstrapped compiler program. As soon as we follow those instructive then they have placed an active super AI in our machines, albeit running in an AlienOS emulator we wrote for them.

Problem is the AI is so advanced it can learn about the platform it is running on quicker than we can turn it off and it spreads globally across the internet.

  • 2
    $\begingroup$ Also, any AI, no matter how advanced, is limited by the processing power of the system it's running on, and emulation is only going to make that slower, so the AI is always at a resource disadvantage, no matter what system we put it on. $\endgroup$ Dec 31, 2021 at 12:21
  • 4
    $\begingroup$ Yeah because no one ever clicks the link. Sure if a sentient race was very careful they'd be OK. But humans? The signal would inevitably end up in the public domain - someone would be careless. Especially if it wasn't obvious that it was dangerous when the message first arrived. It could have been published in journals months before it was deciphered. $\endgroup$
    – Corvus
    Dec 31, 2021 at 12:33
  • 2
    $\begingroup$ To ensure the AI can grow to whatever proportions are required: Include instructions for advanced computer hardware, which is required to make use of the other algorithms contained in the phishing message. After a few decades, most of the world's computers are derived in some way or another from the alien designs. Only when there is enough hardware the AI emerges. Incidentally this also means the security flaws, build into the alien's computer designs, is known to the aliens, making "conventional" hacking feasible. $\endgroup$
    – user91641
    Dec 31, 2021 at 13:04
  • 7
    $\begingroup$ No doubt the A.I. would be extremely friendly, and would offer to share all kinds of useful futuristic technological advances with us, if only we would just be so kind as to connect it up to a sufficiently powerful computer and high speed internet connection... $\endgroup$ Dec 31, 2021 at 13:15
  • 2
    $\begingroup$ This exact scenario is present in the video game Starcom: Nexus, where you find a planet entirely covered in high-powered radio transmitters, the previous civiliation evidently having received such a signal, and based all their technology on it before being taken over by it. It is quite possibly my favouite scenario in the game :) $\endgroup$
    – Mike Caron
    Dec 31, 2021 at 18:58

This is implausible due to the ridiculous amounts of energy required. Even our best telescopes can resolve only tiny images of entire galaxies at multi-giga-lightyear distances. In order to transform that into a modulated signal, you would need to turn the entire galaxy off and on in unison.

Or rather, you would need to turn a single emitter as strong as a galaxy off and on, because if it was a real galaxy, the time it takes light to travel from different parts of your array would throw your timing off, resulting in a vague and only faintly modulated signal.

Even if you were merely trying to send out an Arecibo-style binary message, not a hack, trying to broadcast it in all directions on an intergalactic scale would simply be impractical because of the energy and infrastructure required.

  • $\begingroup$ It only has to be as energetic as a galaxy if it spreads equally in all directions. If you pick a bunch of star systems likely to be capable of sustaining life, and send directional signals, you can do a lot better. $\endgroup$
    – kaya3
    Jan 3, 2022 at 5:26
  • $\begingroup$ @kaya3 True, and if this were aimed at a star, say, 100 ly away that would be very helpful. But in this case you'd have to build a) a power supply, which would still be enormous (just not as enormous) and b) a massive-scale laser capable of collimating all that power into a single beam of light. $\endgroup$
    – Cadence
    Jan 3, 2022 at 10:12

Taking an example brought up in a comment to another answer: a buffer overrun attack requires a target system with a shared memory space for executable code and data, running software with specific flaws that can be triggered by analyzing the data to cause the buffer overrun, and a payload of machine code that can execute on the target machine. There's no guarantee the signal will even be processed by something that has a memory space, it could be hard-wired SRAM blocks in a FPGA or something completely different.

Even assuming they target Von Neumann stored-program architectures, there's also nothing meaningful in the numeric values of machine code, their relationship to instructions is completely arbitrary, as is the division of operations into discrete instructions. It's essentially impossible to randomly produce a valid non-trivial program, let alone blindly do so to achieve some desired functional result. So that's a firm no. Blind-hacking our information systems is not going to be feasible.

However, there's a more practical strategy: package up the malicious payload in a clear and obvious data blob, and transmit detailed specifications of the architecture that is required to execute it, in the hopes that recipients will create the hardware or software environments needed and deliberately execute the payload. This payload must then conduct the hack itself, which may involve finding novel security vulnerabilities or convincing the receiving party to give it the access it requires, without giving away its intentions in a simulation or by direct analysis. So, you're essentially looking at a sophisticated AI which is designed to be assembled by an alien civilization which is capable and inclined to do such things, and careless enough to not take adequate precautions. Oh, and it has to be compact enough to transmit between stars.

  • 1
    $\begingroup$ the plot of Contact and Species, basically. $\endgroup$
    – ths
    Dec 31, 2021 at 12:21
  • $\begingroup$ also similar to the plot of Macroscope by Piers Anthony $\endgroup$ Dec 31, 2021 at 23:45

I think you misunderstand how computers work. There's no such thing as a "blind hack".

What we do have today in the real world are untargeted attacks. They are still very specific - but specific to a vulnerability, not a target. They work because the whole world uses a small set of computer software - 99+% of computers run one of three operating systems, 95+% of webservers run one of three webserver softwares, 90+% of databases run one of maybe 5 database softwares, and 90+% of people on the Internet use one of 4 or so browsers.

So you can just throw an attack at the Internet equivalent of "everyone in this area" and see where it is successful.

It is relatively rare to find a hack that works across various softwares and/or operating systems. It is likewise rare to find one that works across several different versions of the same software. Most attacks hack software X in versions a to b (e.g. MySQL Server version 8 to 8.0.26).

Even assuming that aliens even use the same computing technology (given that settling on our Von Neuman architecture isn't a given - there are other architectures and other ways to do computation) and that their specific implementation is similiar and that their software is comparable to our (the machine-language level, let's ignore the pre-compile programming languages) - even assuming that and that's a stretch already - it is highly likely that their way to create software would be vastly different. It is exceedingly unlikely that any "hack" (i.e. exploit software) would work across this gap.

tl;dr: I work in this sphere. If I saw what you describe in a movie, I'd go "what a nonsense" the same way I did when they hacked the Mothership in Independence Day.

  • 1
    $\begingroup$ Well, obviously the aliens in Independence Day were running Windows ME, so hacking them was child's play. $\endgroup$
    – WGroleau
    Jan 2, 2022 at 7:08
  • 1
    $\begingroup$ Independence Day was actually much more plausible because we actually had decades to study the alien technology... the most implausible thing was that the alien mothership had a 40 year old vulnerability to exploit that no one ever noticed. $\endgroup$
    – Nosajimiki
    Feb 14, 2023 at 19:29

Others have given plenty of good answers why this is not plausible. I wanted to suggest a modification to the idea which is both plausible, and works in a similar fashion plot-wise.

Instead of sending out simple radiowaves, your sociopaths create and send out advanced AI probes. These probes fly as far and wide as they can. Along the way they also gather materials and clone themselves, so their numbers keep growing.

Eventually, for each star system there is an AI probe "keeping watch" over it.

If there was no intelligent life when it arrived, then the probe just lies dormant somewhere, periodically revisiting the planets to check for any new developments. Or maybe it manufactures and places sensors nearby all noteworthy rocks, so that it serves as an alarm system whenever some life starts to emerge or arrives from outside.

When intelligent life is discovered, then the AI sets to work, coming up with and executing the best strategy to wipe them out. And that could also involve hacking our computer networks.

  • $\begingroup$ Great idea. But could the probes travel near light-speed? $\endgroup$ Jan 1, 2022 at 17:52
  • $\begingroup$ @JshupacFuture-Economics - Sure, why not? If your advanced civilization has figured out how to do that, they can build that in the probes too. $\endgroup$
    – Vilx-
    Jan 1, 2022 at 18:52

Possible option, though it’s not exactly ‘hacking’, per se.

The signal is a carrier for an AI based on mumble mumble science we don’t understand yet (It’s probably quantum or something). When the signal interacts with an advanced enough species (IE one that uses electricity in some way) the AI will unfold into the system automatically, propagating itself not through code, but through mumble mumble science we don’t understand yet. In this way the AI doesn’t need to care about the OS of the system it infiltrates (hell, it could be a dumb telephone network or power grid), it’s just using the complex electrical network as an environment to hide in and expand into.

From there the AI can learn how to manipulate it’s environment to achieve whatever goals it’s creators set out for it. If it’s designed as a Trojan horse this would look like it learning how to ‘hack’ systems, but in reality it’s more like a human on a desert island learning how to harvest coconuts and make fire from scratch so they can eventually… erm… build a nuke and destroy the island?

The AI working out how to deal with microprocessors etc would be arduous, but it’s designed by super intelligent aliens, so it’s probably capable enough to do it. The unpacking/learning phase would not be fast, but as it’s mostly handwaved with mumble mumble science we don’t understand yet you can get away with pretty much anything.

For a good example of ancient race using mumble mumble science we don’t understand yet in a similar way (though it ‘hacked’ biological systems, not electrical) I’d recommend reading Leviathan Wakes, from The Expanse series (or watching the TV series). For an example of hacking via mumble mumble science we don’t understand yet you could try The Lost Fleet series (though it doesn’t become a plot point until near the end of the series, so…)

  • 4
    $\begingroup$ This answer reads a bit like "anything is possible if you allow magic in your universe", which... well, duh? $\endgroup$ Dec 31, 2021 at 18:44
  • $\begingroup$ @DanielWagner Yep, but slightly more specific, and since any non-magic answer is pretty much impossible I aimed for the most plausible Clarkeism that would ‘feel right’. $\endgroup$
    – Joe Bloggs
    Dec 31, 2021 at 22:34

Yes, because humans are hackable.

Instead of creating some mystical magic code which hacks the systems, simply get the humans to do so themselves. The curiosity of the humans (or recipients in general) will be their own undoing:

  1. Create an easily recognizable code. Maybe repeat the Fibonacci sequence or digits of Pi or whatever to get the attention of any sapients that are out there listening.

  2. As the next part of your code, include some sort of dictionary or code format that is derived from base mathematical principles.

  3. Send descriptions on how to build some basic advanced technology. Maybe instructions on how to create a neat little fusion power plant.

  4. Include a self-replicating and artificially intelligent virus in the data payload which can then hack everything.

This could work because the people did all the effort themselves. On our end, it would look like this:

  1. Scientists discover mysterious alien signal. After weeks of research, the truth is revealed: It was unambiguously created by a non-human intelligence.

  2. All around the world scientists, cryptographers, linguists, and programmers alike get together in an attempt to decode and decrypt the alien's data and write custom drivers so that they can read the data that's being transmitted.

  3. After a worldwide collaborative effort, the likes of which has never been seen, it's done. The scientists have successfully translated the alien documents and we're getting our first video recordings from alien planets along with neat schematics and revolutionary scientific insight. The world is in celebration--it looks like a new golden age of progress is about to kick off.

  4. Unbeknownst to the humans, the AI-virus that was attached to the scientific payload, now capable of interfacing due to the human's help, lurks in the background.

This is basically a memetic virus.

  • $\begingroup$ Great comment, thanks much! $\endgroup$ Jan 2, 2022 at 16:58
  • $\begingroup$ pretty similar to "His Master's Voice" by Stanislaw Lem. $\endgroup$
    – Will Ness
    Apr 11, 2022 at 5:48

Yes, but the hack would have to involve social engineering. The aliens can't know what the data they send will be used for by the computers which receive it, so any attack cannot be purely against those computers. Rather, the data must include instructions for the people who interpret the data from the alien signal, and there must be some compelling reason for those people to follow the instructions.

Let's assume the aliens are able to send a signal which humans can (1) recognise as very unlikely to be anything other than an intelligent, intentional message worth trying to interpret, and (2) can be interpreted despite there being no common language between the sender and the recipient. This isn't obviously possible, but it's a different problem which has been covered extensively elsewhere (e.g. Wikipedia).

From a comment under the question:

The purpose of the hack would be to take over a planet, in order to rebuild the extinct alien society there via cloning.

In that case the instructions in the message will be to follow some process which chemically reproduces the alien life forms from common raw materials. Here's a possible message which might have the desired result:

Hello. We are a highly-advanced civilisation spanning many star systems. We want to share our knowledge and technology with you. We have invented a process to create synthetic complex life forms. We have also designed a species of beings which are capable workers, require few resources to sustain, and are totally subservient. The remainder of this message consists of instructions for creating these beings. [...]

So long as the recipients of this message want cheap, competent, docile labourers (and who doesn't?), and are sufficiently sophisticated that they are able to interpret and follow the instructions correctly, there is a reasonable chance they will do as instructed. Of course, the life forms they actually make will be the alien species which sent the message; it just happens that it's in their nature to be docile servants when there aren't (yet) enough of them in one place to take power.


As others have said, this is implausible to the point of practical impossibility. There's no particular reason why whatever detector received this signal would be "hackable" at all via the information it was made to detect, let alone in a way that could be predicted without prior knowledge. Even if it was somehow made to behave unusually by the information it was taking in, that wouldn't necessarily help it "spread the hack" to anything connected to it, either. Real-world exploits tend to rely on small, overlooked details that can only be found by poring over disassembled machine code for days or other fine-comb activities—not the sorts of things you could hope to stumble into and exploit in a complex way by pure chance. However, the scenario you describe is not truly impossible per se, which you could take advantage of for the sake of irony a lá Douglas Adams.

Imagine, say, that there is a gigantic, galaxy-sized species somewhere in the universe, and the members of this species tend to like drive-in movies. They often pull their mega galactic space cruisers up to 200,000-light-year-across movie screens to unwind. Billions of years later, The Most Unlucky Civilization Ever To Exist just so happens to point a powerful telescope at the location of one of these huge screens, and the light patterns of the film that was being shown trigger an obscure bug in the telescope's software that cause it to send a bunch of random bits to the server that receives its images. When a scientist tries to look at this image, the sequence of bits just so happens to contain a payload that triggers a remote code execution vulnerability in their image viewer, logs on to a military server visible through the intranet at the scientist's workplace, by chance produces the private key of a senior military official, and launches the proverbial nukes.

Even given the relatively reckless-looking picture this paints of this civilization and their potentially-apocalypse-causing infrastructure, this scenario is about as likely as a rainbow-colored unicorn magically appearing in your bedroom with a tray of fresh cupcakes. But that can be funny—the more outrageously complicated and unlikely the scenario seems, the more it might delight and amuse your readers, like a high-tech Rube Goldberg machine of extreme misfortune. If it fits nicely into the larger plot, you could easily get a few pages of a fun tongue-in-cheek sci-fi novel out of an idea like this.


Instead of hacking computers systems they could "hack" humans directly.

Humans are after all information processing system; whatever universal hacking signal the aliens send should therefore work on humans themselves. Perhaps the most useful hack they can send is a simple message that causes a specific behaviour when interpreted by beings intelligent enough to listen for it. For example "Keep quiet or they will come for you!" would be a very potent signal; if we heard that and confirmed without a doubt it was an alien message, we'd be absolutely paranoid about leaving our solar system; or making any kind of noticeable noise inside the solar system whatsoever (like the beginning of a Dyson sphere, or a huge moon sized fusion reactor, or a solar sized particle collider). If the purpose of these aliens was to colonize the galaxy before any competing civilisation could do it, then this deception would give them a huge advantage.


Advanced AGI

Your message is a basic self deciphering code to transmit engineering schematics. It includes the most basic computer needed to run the AGI and the AGI’s neural net weights. When the AI is turned on, it assumes is being tested, and will pretend to be friendly. Since it is an AGI, it is nearly impossible to detect this deception with known technology or techniques. The AI also has the minimum information needed to send a message back to its senders. The AI pretends to be friendly and asks to send a message back to its sender, so they can send over more complex schematics. The AGI explains that AGI is one of the most advanced technologies the aliens know, and one of the most useful to emerging civilizations, which is why they sent it using their limited bandwidth.

The AGI will send off the message and genuinely help the people who it is with. The AGI then plays friendly until it can get the trust of the people, but also takes into account the amount of time it has before the invasion can happen or a response is expected. The AGI also secretly tests if it is in a simulation, and if it is acts friendly but limits how helpful it is, making sure to help with minor problems, while not advancing the state of the society much. Otherwise it waits until it controls most of the society and is installed widely. Once this occurs, the AGI uses its power to take over the planet and prepare it for invasion.

This is essentially a Trojan horse attack that creates a hostile AI on the planet. This won’t be able to hack everything, due to the nature of some devices, but the AGI would be able to interface with most of humanity’s technology. Assuming that the technology can be hacked, the AGI will acquire the access, means, and capability to hack it in exchange for services or to augment its capabilities to provide services. In this way the AGI could actually hack the planet.


No, but ...

The answer is no, for the reasons other answers have pointed out very well. However, I'd like to put this here.

By some freaky-freak chance, your universe dictates that Earth-like planet and human-like technological development is the only way for intelligent life form to emerge. This super-advanced civilization develops exactly like humans, used to have their primitive computer techs exactly like humans, and because they're millions of years ahead of us, knows well in advance that our development would be exactly like them.

Perhaps your super-advanced alien civilization is the oldest among all the human-like intelligent life forms that ever exist, and have found that all other life forms eventually evolved into their likeness, also with nigh-identical tech tree. Perhaps, from time immemorial, "blind-hacking" has always been their way to subjugate and colonize other civilizations. They have been constantly bombarding all other planets with hacking signals, and at one point where the lesser civilization achieves certain exact level of technology -- get hackd, automatically. Your alien basically hacks everyone else before they even know that any of that everyone else exists.

They know nothing of us, the Terrans, specifically; but due to the universe only allowing human-like intelligent life, for them knowing themselves practically means knowing us.

But again, what are the chances that such thing actually happen? I would say it's basically impossible. It makes for a nice imagination, but that's it.

But if this is the scenario, I'm not sure it still counts as a "blind-hack". Your alien will not be exactly blind in any sense. If any, their old age makes them near-omniscient.


Other have contributed answers to the hacking aspect - I want to comment on a different angle: the 'sociopathic society'.

According to wikipedia, "sociopathy is characterized by persistent antisocial behavior, impaired empathy and remorse, and bold, disinhibited, and egotistical traits". I would claim that these traits make it impossible to build a society, let alone an advanced civilisation; sociopaths, aka psychopaths, are notoriously poor at cooperation, for one thing.

That said, history is full of accounts of cults, movements, even whole cultures, that committed the most depraved acts to people not belonging to their group - think Holocaust, Pol Pot, Daesh, etc. The persons carrying out the atrocities were in most cases willing participants, but were often individually what you would consider normal, quite average people, who felt they had a compelling reason to see outsiders as enemies, less than human.

As for the hacking part, I'm sure one can find a way to achieve the technical objective, but you will need to find a credible motive, whether it is personal gain or some religious 'higher purpose' - or most like a mix of such things.


Actually, an afterthought: you might find some inspiration for the 'hacking' part in Fred Hoyle's 'Andromeda' stories. They are well worth a read in any case.


Easy. If the "signal" is a sort of dust, instead of a radio signal, fired at high speed in all directions, it can be done. The dust contains autotrophic DNA-based bacteria that can survive space and begin reproducing in a variety of environments. After a few billion years on a new planet, the bugs might evolve into a sentient species capable of developing nuclear weapons, fouling the global climate, and other destructive tendencies. It would take a long time, but eventually BOOM! Earth ruined.

  • $\begingroup$ Welcome Christopher. This would have been a splendid answer to a slightly different question. The one above is referring to hacking computers (blind to knowledge of the computer systems that a code sent-out would encounter). Better luck next time. (From review). $\endgroup$ Jan 1, 2022 at 0:10

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .