2
$\begingroup$

In the near-future (circa 2030), publicly available open-source A.I. can create any Picture, moving picture, or audio track imaginable from a text prompt. It takes a lot of computing resources (but there's quantum computers) but it's not beyond someone's means to source such computer CPU time.

Think of today's OpenAI "Dall-E 2" or Google "Imagen" or Hatsune Miku to the power of ten.

For example,

Get me a "rule 34" undercover home video of Celebrity+A and Celebrity+B first nuptial night.

No matter who those two celebrities were (or even if they were actually married), there you go, a few moments later, you have your pixel perfect, photorealistic, 400K resolution undercover video. Even the moles and creases, complete with pitch-perfect audio tracks with ambient sound (crickets, cars, neighbors snoring and chattering, TV in the background, creaking floorboards, etc...), voices made by vocaloid-like software and sound effects generated from scratch that are identical to real ones.

Quantum computers made encryption a joke. Cryptographic algorithms used in the 2020s can be cracked in microseconds.

This is the scenario.

Our (2020s) legal system often uses photographic and audio evidence. The Jurors love a good video, they say.

Ten years from now, the videos floating around are just too good, to say the least.

However, when you can just make convincing evidence that can fool every single flesh being and almost any other software, how would the court accept evidence as reliable?

Assume analyzing forgeries (i.e. A.I.-created media) is too expensive for any but the most high-profile cases as you need to use even more powerful AI to check the media, and it is not fool-proof (not to mention hard to cross-examine. You'd need a third AI and then it's the "turtles all the way down" dilemma).

Improve this question
$\endgroup$
10
  • 1
    $\begingroup$ "Cryptographic algorithms used in the 2020s can be cracked in microseconds": No, they cannot. Plain ordinary widely used AES-256 is not afraid of quantum computers. (And it's not the only one. In fact, we already know what we need to do to make HTTPS and SSH impervious to quantum computing attacks, and the relevant standardization bodies are actively pushing for it.) $\endgroup$
    – AlexP
    Jun 13, 2022 at 20:10
  • $\begingroup$ I agree with you but I want to make two objections. 1) people often use passwords with less than 256 bits of entropy (making cracking the PW easier than the algorithm) and 2) the premise of the question assumes [future-tech] has QC that can do that. A quick search showed that it can be done, with enough (mind-boggling lots by our current standards) Q-bits. $\endgroup$
    – Mindwin Remember Monica
    Jun 13, 2022 at 20:54
  • $\begingroup$ Deep fake has been used against celebrities in smear campaigns. But they still got caught and sued. $\endgroup$
    – Faito Dayo
    Jun 13, 2022 at 20:55
  • 5
    $\begingroup$ I’m voting to close because: (a) This is an off-topic high concept question (hypothetical, opinion-based). (b) Questions about how to stop perfect things from being perfect violate the "all answers are equal" rule. (c) Entire libraries exist concerning the law (local, state, national, international...). Even if the exact and specific jurisdiction were provided, this violates the help center's book rule. $\endgroup$
    – JBH
    Jun 13, 2022 at 20:59
  • 1
    $\begingroup$ @AlexP Your assertion that video (like the audio of yesteryear) would simply become untrustworthy (thus trivially solving the law's problem) is highly likely. What would be a better problem is realistic holography that could be projected into a cluttered space without local projectors or distortion. Let's add self-driving cars. Now you have living witnesses watching Anne Hathaway steal a car. But one would still have all the tried and true methods: fingerprints, DNA, proof of her actual itinerary. Your comments have led me to wonder if the law would care at all. $\endgroup$
    – JBH
    Jun 14, 2022 at 1:55

4 Answers 4

Reset to default
9
$\begingroup$

  1. Justice worked just fine before the advent of mass video and audio surveillance.

  2. Not hard to make video cameras and audio recorders digitally sign the video and audio clips. Quantum-computing-proof digital signature algorithms are already available.

  3. We already have the means to make just about perfect still photography and audio fakes. This is why judges insist on establishing a chain of custody.

Improve this answer
$\endgroup$
1
  • 2
    $\begingroup$ For some definitions of the term "just fine." $\endgroup$
    – Mary
    Jun 14, 2022 at 0:17
5
$\begingroup$

Theoretically, technology that can spot fakes would also improve.

While some people are making technology that can create realistic fakes, there will simultaneously be people who are working on complex ways to see through these hyper-realistic fakes.

Though it is hard to tell a fake photo from the real one at first, there will be technologies out there that can distinguish between the two that theoretically would allow the public to know the difference.

However, that doesn't solve the whole problem.

The availability of this technology becomes the real issue. Just because means of verification exist does not mean that people will use them.

If everyone and their grandmother can make and spread false videos that seem incredibly realistic, there will be people who fully believe it no matter how many times that professionals verify it is fake. This is how conspiracy theories form.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Unfortunately, as AI-generated text becomes better, there will not be a way to distinguish them from human-generated words. arxiv.org/abs/2303.11156. It could be that this mathematical relationship will apply to images and videos, as well. $\endgroup$
    – chase leffers
    Apr 30 at 14:35
4
$\begingroup$

Dystopian Surveillance

enter image description here

Your world is a scary dystopian future where computers decide what is true and what is false. Let's take it to the extreme. Your universe also has round the clock non consensual surveillance of its citizens.

The evil machine can create an hour long hyperrealistic video of Martin Clunes and Geena Davis cheating on their respective partners with each other.

enter image description here

But this is not enough. The court will simply review the ubiquitous footage of Martin and Geena and spot they were nowhere near each other at the time. Martin was at the zoo with his grandchildren. Geena was eating a cheeseburger in the bath.

To be convincing the machine must delete and fake a section of footage around the event. It has to fit the hour long video into the security footage of them entering the building, leaving, and what they did before and after, perhaps for days or weeks leading up to the affair.

It takes longer than a week to create a week of fake security footage. And of course you need your fake week to not contradict other events that happen during the real week. Since the real week has not happened yet, the fake week is hard to achieve.

Improve this answer
$\endgroup$
2
$\begingroup$

Falsified evidence is a problem as old as law enforcement, so this is not a new problem.

One way to discredit fake evidence is by putting the chain of custody into question. What were the circumstances under which the evidence was created? How was it obtained? Who can attest to that? Who had access to the evidence before it came to court? Is it guaranteed that nobody else had access? Does anyone in that chain of custody have reasons to falsify evidence? If no, then the evidence is probably reliable. This is actually already very common in our legal system of today. There are lots of forms of physical evidence which were very conclusive if true, but could be easily manufactured. Like documents (physical or electronic) or DNA samples. It's not uncommon that lots of potentially incriminating evidence is thrown out of court because the chain of custody is put into question.

Another way to prove evidence as fake is by looking at inconsistencies caused because who (or what) created the fake had to improvise on information it hadn't available. For example, when the fake sex tape shows physical characteristics of Celebrity A which don't match their actual body (yes, there are nude scenes of them from movies and modeling, but the public does not know that those also use AI technology to change certain features of their body). Or because it shows details of the filming location which don't match real-life.

Also keep in mind that AI technology can also help the legal system. For example by offering more reliable lie detection. The current lie detection technology (the polygraph) is known to be very unreliable. AI technology could be used to create far more reliable lie detection and use it to verify witness testimonies with higher accuracy.

Improve this answer
$\endgroup$

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