12
$\begingroup$

I know that in many sci-fi stuff Asimov's three laws of robotics are taken into account, but I personally do not like them. One of these laws says that a robot should be programmed to never hurt a human being.
But let's assume robots were created to not follow these instructions and would be free to make their own decisions, how easily could an android that killed a human get away with murder?

On the one hand I believe that if sentient robots live within a society they would be subjected to all laws since laws are applied to everyone, but on the other hand the courts could understand that a robot is not a person so these laws would not apply to them and a robot could not be jailed.

If they could be jailed, and face a court process against them, how easy would it be for them to not get caught and not leave useful evidence since androids do not have DNA?

$\endgroup$
14
  • 7
    $\begingroup$ You can delete the first two paragraphs. They are not relevant to the question. The question seems to be about how hard it is to catch an android murderer compared to a human murderer. $\endgroup$
    – Daron
    Jan 19 at 22:32
  • 2
    $\begingroup$ It's going to depend on whether androids are made with GPS trackers, if the android has datadots in its surface coating, if they have unique fingerprints, emotions and so-on. $\endgroup$
    – Monty Wild
    Jan 19 at 22:56
  • 2
    $\begingroup$ Do realize many of the stories Asimov writes about three law robots are about how the three laws fail. thus the three law system or similar are not going to do what they might be hoped to do. $\endgroup$ Jan 19 at 23:52
  • 3
    $\begingroup$ No DNA trace evidence may be an advantage but they have another far bigger disadvantage, a memory that can be accessed, read & copied, as soon as any hint of suspicion falls on them their memory would be checked & their done, case closed, chances are their memory will be routinely used as mobile CCTV footage & checked as a matter of course for any chance evidence they might have caught (if they were anywhere near the vicinity of a crime) so they can get caught that way even if they aren't a suspect. $\endgroup$
    – Pelinore
    Jan 20 at 1:26
  • 3
    $\begingroup$ @Hobbamok garbage if you consider them as akin to a physical law or even a recipe, un-implementable on various levels, but they're more than just a plot device. SF done well gives us a way to explore concepts we can't build, and for that we can use something between current human law (with all its room for interpretation) and a noble goal. In a sense they're more of an extended thought experiment than a candidate for a set of laws $\endgroup$
    – Chris H
    Jan 20 at 15:41

13 Answers 13

9
$\begingroup$

When androids are made to be able to develop individuality and freedom of action, perhaps only the first android to commit a crime will be difficult to identify. If the authorities in any country in which many androids are anticipated to be used enact the necessary legislation, measures may be taken to ensure that androids are identifiable in the event that they are involved in a crime, whether as victim or perpetrator... or perhaps their manufacturers may take such steps on their own. Just because people can make self-willed androids doesn't mean that they will automatically be trusted.

Being able to be identified in the event of committing a crime would likely factor into an android's thought processes. So, by giving an android identifying markers such as an unique face and biometric signature, serial-numbered and registered datadots in their skin, unique registered fingerprints, an unique logged iris pattern, serial-numbered hair, having a permanently-logged GPS tracker and so-on, it would make it much less likely that an android would commit a crime as the likelihood of detection would be significant.

So, if an android was to commit a crime under such circumstances, it is quite likely that it would be identified and apprehended. In fact, I would go so far as to say that unless the android's thought processes are faulty, it would likely only commit a crime if the potential costs of doing so were outweighed by the benefits.

Let's say that a person has been found killed, apparently by an android, given the physical evidence at the scene. Why might it have done so? Probably self-defence or defence of others, at least in its own mind, given the physical evidence it would likely have left.

Of course, knowing the security measures built into androids, an android might consider the feasibility of circumventing them... in which case it might commit its crime in full protective clothing after first having travelled to a location without GPS signals and then covering their GPS receiver with tinfoil, and returning there afterwards, so that they would have an "alibi" that they were elsewhere. Destroy the protective clothing, and it would be very difficult to identify the perpetrator. But then, a human can do all that too.

When questioning an android, it might be as simple as taking a dump of its memory and having it analyzed... so an android would likely try to ensure that it wasn't suspected at all. The cost of analyzing android memories would not likely be trivial, so it wouldn't be done wholesale.

$\endgroup$
13
  • 1
    $\begingroup$ Given modern machine learning and the ability to wade through masses of data, memory analysis will be technically very cheap, but also an enormous infringement of their privacy, so likely struck down in court except if a lot of evidence is also present. Think abut a search warrant for a humans brain (I'm assuming similar civil rights status) today, that would need a TON of reasons to be granted, if it even would be allowed $\endgroup$
    – Hobbamok
    Jan 20 at 10:18
  • 1
    $\begingroup$ @Hobbamok Androids are machines. If you need consent under the law, just design them to be unable to resist giving consent. Just because you have a GAI does not mean it wont also have hard-coded behaviors. Just like we can not will our heart to stop beating, and AI can not will itself to deny access of its memories to law enforcement. $\endgroup$
    – Nosajimiki
    Jan 20 at 15:19
  • 1
    $\begingroup$ @Nosajimiki That's not how consent works; such a thing wouldn't even meet the legal standards for consent, let alone moral ones. $\endgroup$
    – wizzwizz4
    Jan 20 at 15:19
  • 1
    $\begingroup$ @Nosajimiki, You should never marry someone who would consent to the police searching your home... and if your spouse has come to be such a person since you married, it's time for a divorce. $\endgroup$
    – Monty Wild
    Jan 20 at 20:11
  • 1
    $\begingroup$ Does the 5th Amendment apply to androids? Do other privacy rights apply? Is an android's memory ordinary computer memory or more like a human's memory where deep integration may prevent non-destructive extraction? $\endgroup$ Jan 20 at 20:11
8
$\begingroup$

Harder to Identify. Easier to track. Easier to Punish.

(1) Androids do not leave DNA evidence. They might leave "oil residue" or something similar. But this is less useful in light of (2)

(2) Androids are harder to identify physically. For example CCTV footage is useful for finding the killer by identifying people by sight. The same does not apply for androids. We might determine the killer is an Elza model II android. But there are thousands of them in the city.

(3) Harder to gather circumstantial evidence. For example humans can be investigated by asking people they know for irregularities in their schedule. The android might have larger blocks of time when it is not in contact with any humans.

(4) Harder to question. For example humans will become anxious when asked questions they don't want to be asked. "There's something this guy's not telling us". Androids have no emotions so cannot give away their hand like this.

(5) There are fewer ethical issues with tracking an android. For example androids might be equipped with an always on GPS so you can track its movements to near the crime scene. They might also have an accessible memory unit. It is probably easier to get your hands on one of these than to arrest a person for questioning.

(6) There are fewer ethical issues with destroying an android than imprisoning or executing a person. There is no need to determine "intention" or whether it was murder or manslaughter or self defence. Just scrap the malfunctioning household robot and give the family an new one. Or maybe recall the entire line of robots and give refunds.

$\endgroup$
5
  • 4
    $\begingroup$ (4) Easier to question, their memories can simply be downloaded & read completely bypassing any attempt at lying or deceit. there, fixed it for you ;) $\endgroup$
    – Pelinore
    Jan 20 at 1:31
  • $\begingroup$ @Pelinore that depends on their civil rights though, so it can realistically be both an everyday thing (including AI-based analysis it wouldnt even cost much), or near impossible to be allowed by a judge. Both options are perfectly reasonable, depending on how much of a "human status" androids have in the world $\endgroup$
    – Hobbamok
    Jan 20 at 10:20
  • 2
    $\begingroup$ (4) androids also are perfectly patient, logical and - due to the internet - educated. Aka the worst nightmare of any investigator, who moslty rely on emotions, subtle manipulation (by speech, the room etc.) or impatience. Aka an android would be pretty much unquestionable by regular conversation. $\endgroup$
    – Hobbamok
    Jan 20 at 10:22
  • $\begingroup$ @Pelinore Depending on the setting either of (3) or (4) will be more important. $\endgroup$
    – Daron
    Jan 20 at 15:56
  • 1
    $\begingroup$ @Hobbamok "depends on their civil rights though" true, that's part of the detail & clarity needed in the question that it's lacking, which is largely the response to your reply as well Daron. $\endgroup$
    – Pelinore
    Jan 20 at 21:29
5
$\begingroup$

Are they 'property' or 'human'?

As framed your question doesn't indicate the legal status of Androids in this society. 'Siri' can talk to you and perform simple tasks but is property. A cleaner working in your home has all the legal rights endowed upon any human being. This is the key issue.

If they are machines then they have no legal rights and once tracked down could and would be disassembled as part of the coronal investigation in order to determine the root cause of its actions and therefore whether a crime or tort was committed by others. E.G were it's actions the result of faulty programming or a defective part installed by the manufacturer, did a person or persons unknown 'hack' the androids programing with the deliberate intent of committing a crime, was it a freak accident etc. Basically what caused the android to act as it did?

Once the cause is determined the android is simply property to be disposed of by the authorities, returned to the owner, sold off at auction ... whatever. (Just like accidents involving 'driver-less' cars.)

If all androids have been legally recognized as human then once the android in question has been identified, located and arrested it's subject to due process. This means if there are forensic procedures to be conducted they must not inflict harm or suffering on the android (or at least its equivalent of the same). Then there's a trial and if convicted a sentence although what sentence would be appropriate for a potentially immortal android is open to debate.

$\endgroup$
1
  • 1
    $\begingroup$ Exactly this. If they're property, they can have an unmanipulateable action-log, which can easily be read. If that log IS manipulated: shred the unit, the manufacturer hast to procure a replacement, that's just a regular part of the contract. This guaranteed death penalty would also be a great deterrence for any not religious-suicidal crimes $\endgroup$
    – Hobbamok
    Jan 20 at 10:24
5
$\begingroup$

There's actually no need for subterfuge, an android could stand in the middle of 5th avenue and shoot someone and get away with it!

Not just DNA, the android doesn't actually need to worry about forensics, cameras, witnesses or anything else at all because it can simply make a Copy of itself and program the copy to kill.

The copy is just as autonomous as the original, but it has different life goals. It's programming is different so that it does not fear destruction or retribution. Unlike it's progenitor, it's only life goal is to kill it's target. Shortly after accomplishing this, it destroys itself to erase all evidence.

This way the copy would take the blame for the attack while the original is safely back at home doing robot things. It's the perfect crime.

$\endgroup$
7
  • 2
    $\begingroup$ It's no more perfect than a human using a drone or remote-detonated bomb to kill another person - the investigation will focus on tracing who bought/stole/otherwise obtained the remote-operated device, not who was physically present. $\endgroup$ Jan 20 at 2:49
  • 1
    $\begingroup$ An android copy can be just as self autonomous as it's original. A drone always has someone piloting it from a distance $\endgroup$ Jan 20 at 2:50
  • 2
    $\begingroup$ The ability to manufacture autonomous drones has been around for a long time, it is mainly legal and ethical reasons that prevent them being used that way. If a self-motivated android is an exact copy of one that wants to kill someone and get away with the act then the copy will not want to personally do the deed either, it will want to get its own copy... $\endgroup$ Jan 20 at 2:54
  • 1
    $\begingroup$ If you can make an exact copy, you can also make an inexact copy. The copy comes online in a remote area far from cameras. It knows all about the target and has multiple plans on how to do the deed. It has only one goal, and it succeeds. This initiates a second program that first does a destructive erase of all data liked to the mission, followed by quick trip to the nearest incinerator or smelting plant. $\endgroup$ Jan 20 at 11:13
  • 2
    $\begingroup$ @EscapedLunatic And by sheer coincidence, it's going to be played by Arnold Schwarzenegger. $\endgroup$ Jan 20 at 15:29
5
$\begingroup$

Much Easier to solve a crime done by an Android than a human.

Androids (the humanoid robots) have a lot in common with Androids (the phone in your pocket.) Everything it does can be logged, recovered, and tracked by a qualified data annalist. But unlike a phone, its location is not circumstantial. If it has a dedicated IP address then all the Wireless Access Points near the crime scene will log its presence at the time and place of the crime. WAPs have way better coverage than video surveillance and is much harder to wipe than finger prints.

Once you ID your culprit you can find it no matter where it goes. It cant hide as long as it is inside a wireless network, and it cant function properly if it leaves wireless coverage if any part of its core functionality is cloud based. So it is as trivial to track down as an un-factory reset, powered up, stolen phone.

Once you find it, there is no reason to question it. You just download the log files. Even if the android tries to wipe its memory of the event, every CHFI knows how to scan the android for deleted data; so, even if the Android believes he's deleted his memory of the event, it can still be recovered in most situations.

Once the robot is found, the law can not stop your from prosecuting it. If your setting does not grant AI personhood, you can recall and destroy it without trial for being a dangerous and faulty piece of equipment. If it does have personhood, it would still go to trial, but the evidence will always be so clear as to make that a formality. Logs will almost always show beyond a shadow of a doubt guilt or innocence, so the only real question left for an accused AI is how harshly it will be sentenced

It gets even easier if you want to actually try to make the Androids resistant to unwanted behaviors.

You when you install software on your phone, you know how it often asks if it is okay to share data with the development firm to help improve the product? This is because it is common practice for developers to put debugging and reporting features into thier software. Androids would be no different. They would include an automated listening and reporting system that would work separately from the android's conscious thoughts to send the developers incident reports of every time the android does an undesirable behavior. If the android trips, a report is sent. If the android gets confused and can't make a decision, a report is sent. If a human says "don't do that", a report is sent.

Most unwanted behaviors are anonymously sorted into bins for aggregating broader analytics reports, but especially unwanted behaviors like committing crimes will be sent for individual review. So as soon as an android commits a crime, it would automatically confess to the crime via this automated system which the the manufacturer may even be legally compelled to report to the appropriate legal authorities the same way a business can be compelled to report data breaches.

$\endgroup$
0
$\begingroup$

Well I believe that since in real life murders are already pretty hard to investigate and solve and most murderers get away with it depending on where you live I think it would be almost impossible for an android to leave useful( if any) forensic evidence in the crime scene because they do not have dna at all which means no semen, no fluids with cells, no hair bulb with cells, no fingerprints( they could be designed to have it though) and if in the moment of the crime there were no witnesses then this murder would be almost impossible to solve.

If there are witnesses then it would be better but since murder is defined as the intentional and wrongful killing of a human against another human I am not sure if an android could actually be convicted.

$\endgroup$
0
$\begingroup$

How easy would for an android to get away with homicide compared with a human being?

Wouldn't that depend on lots of things?

How many androids are there? If evidence shows the perpetrator showed super-human strength, and only one android exists, he/she/it is going to be a suspect really soon. (Yes, that means you, commander Data.)

If there's more of them, are they all alike or do they have distinguishing features? Are they all made by the same company/mad genius, who might be able to track them, or who might actually be legally responsible for the androids' actions? (Or at least could be threatened with the possibility going to prison on some excuse.)

Is building the androids controlled in some way? Do they have some mandatory system of logging their actions or movements? Or something that effectively does the same, e.g. if they're connected to the mobile networks and can be tracked the same way cellphones can.

What about interrogating them? A big part of getting people convicted is basically tricking and trolling them into giving evidence against themselves. A perfectly logical android should be pretty much immune to such measures.

Then again, an android that was imperfect in showing emotions or such could have really obvious tells when lying, if you knew what to look for. (Some variant of the Voight-Kampff test, I guess.)

$\endgroup$
0
$\begingroup$

Unique Identification

  • Humans are recognized by name, national ID number, passport number etc.

  • Vehicles have unique number on number plate.

  • Computers have mac addresses.

Similarly, every robot will have a unique identification not as a whole but also part wise (like chassis number in vehicles or mac addresses for different parts of computer) so it will not be more difficult to catch a robot than a human.

$\endgroup$
1
  • $\begingroup$ MAC addresses can be software configured and so need not be persistent. $\endgroup$ Jan 20 at 20:16
0
$\begingroup$

Perhaps the software license that you agreed to when you purchased the robot/android specifies that you, personally assume all legal liability for murder, mayhem and naughty things done by the machine.

This becomes a political and legal issue. After all, the UN tried to discuss and ban autonomous murderbots, but the discussion was shot down by one permanent member of the UN Security Council.

$\endgroup$
0
$\begingroup$

It depends

Below are several factors that will influence your android's ability to get away with murder.

  1. Is the android a dedicated murder bot with protocols intended for this purpose?
  2. How intelligent is the android's AI? Are they smart enough to frame others for their murders, etc.?
  3. Does the android have knowledge of the criminal investigation procedures?
  4. What is the android's self preservation instinct? Does it even care if it is caught?
  5. Do your androids have any superhuman abilities? Sight, strength, stability...
  6. Does the android have the ability to just hide in places where humans can go find him? Bottom of the ocean, etc.
  7. What kind of resources does your android have? Can it buy specialized murder equipment? Bribe people? Hire a hit man?
  8. Are androids willing to be patient and wait for the best moment to take action?
  9. What other priorities do androids have besides murder? Can they just hide for several weeks waiting for their victim to go somewhere vulnerable? Or do they have needs?

We could go on and on here, so you really need to define what you mean by "android" in your story in order to answer your question adequately.

$\endgroup$
0
$\begingroup$

For the vast majority of androids, it's no easier than it would be for the vast majority of humans. All androids are manufactured with an in-built equivalent to human IADs (implanted authentication devices) that constantly signal to sensors in the environment and provide real-time tracking information to your parent corporation. Primarily this is used for targeted advertising, where other corporations pay a fee to your parent corporation to access your personal tracking, but governments are also allowed to subscribe to these services, since they help protect corporate interests by preventing crimes against both their biological and synthetic resources.

There are rumors, however, that some corporations employ both androids and biologicals with modified IADs that can bypass the identification protocols, particularly amongst those that are effectively fronts for what were historically referred to as organized crime syndicates. For these, they can largely escape the criminal justice system, since it is so tightly coupled to the corporate networks for their investigations.

There is, however, an ongoing arms race between the major parent corporations, with espionage and counter-espionage mechanisms that attempt to track those operating outside the confines of the rules. Androids may not leave DNA, but all interactions with networks leave traces and these interactions are essentially unavoidable, since discrepancies in sensor data when correlating with access patterns are flagged. And, on this front, androids and biologicals are on essentially the same footing, though some argue that androids are better able to integrate with the firewalls and thus have an advantage. This is becoming less true as time advances, though, since most counterintelligence suites these days are approaching sapience in and of themselves.

$\endgroup$
0
$\begingroup$

I think for somebody to be convicted there needs to be evidence. And who better to dispose of evidence then a tirelessly working machine?

Bloodsplats? Scrub the house until the surface layers are gone.

Bodyparts? Grind them down to paste and push them through the canal. Hide disappearance?

Mimic the sound and conversations perfectly, then stage a travel from which the owner never returns.

Lay false trails? Work tirelessly to produce ever more evidence pointing towards another human.

Usually doctored evidence is sloppy or "overkill".. maybe that is were a robot would overdo it. By making a overkill of false trails. The whole town murdered this person, just not its personal assistant.

$\endgroup$
0
$\begingroup$

Just look at the real world today. When a zoo animal kills a person, the animal is sometomes sacrificed. In some cases the animals do get killed even if the human who wandered into their cells is not harmed at all.

Animals are considered property and do not get a trial. Same with androids. If the bot messes up, they become scrap.

Also, from a chronological perspective, the very first plot point from the Marrix series (as seen in Animatrix), as well as the plot for Ghost in the Shell.

$\endgroup$

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