There’s only been a short period of human history from the invention of photography to today. We had evidence before photography existed, and we will still have evidence even if photography can be trivially faked.
There’s only been a very short period of human history where video cameras were cheap enough to be used for widespread surveillance, but could not be trivially faked. That period is just about over. We had laws prior to video surveillance, and we will still have laws even if video surveillance becomes obsolete.
But it won’t. Instead, provenance, or chain-of-custody of evidence, will become even more important.
You can fake security camera footage — but if real security cameras upload their recordings automatically to a service that timestamps them and certifies them, then that metadata (and the trustworthiness of the service) represent a way of verifying that particular footage was created at a particular time, and even by a particular instrument.
Instead of Joe’s Corner Store having video cameras that record only to local storage or to Joe’s own account on a cloud service, they will instead stream to a service run by a security or insurance company, or (in some places) the police. This service will timestamp the video, record checksums, and thereby provide assurance that a particular video recording is really from Joe’s camera and not faked by AI.
Effectively, you can’t trust a mere video that appears to show Taylor Swift shoplifting from Joe’s Corner Store — but when a representative from Joe’s insurance company testifies in court that the video was definitely recorded by their device at a particular time, and has the logs and checksums to prove it, Ms. Swift will be in trouble.
In the Court of law you are correct
In the court of public opinion… yeah we had better figure out how to quickly prove one from the other or a lot of people are going to have a very rough time
Nah.
Even if it were technically possible to conclude a video is “genuine” (whatever that means) the genuine public would not trust it. As in… vaccines method of action is too complex to understand, therefore facebook memes are a more reliable source of factual information.
I’m hoping that fake video becomes so prevalent that absolutely everyone is forced to acknowledge that no video can be trusted as a source of factual information.
This was once a fake photo that fooled many people. People figured out later that fake photos are a thing, and now only the most gullible are fooled by Bat-Boy or alien autopsy “photos”.
No need to stream the whole video externally, you could just send the checksums every X minutes and then provide the video with that checksum later.
It doesn’t entirely stop the problem though, as you could still insert faked videos into the stream. You just couldn’t do it retroactively.
Sure, but robbing a store and simultaneously hacking their video feed is harder than robbing a store and retroactively creating fake footage.
I’m not particularly worried about robbery. There are far more sophisticated ways to attack an organization or person.
My prediction: verified video will start to become a thing.
Phones will be able to encode a digital signature with a video that certifies the date, time, and location where the video was captured. Modifying the video in any way will invalidate it.
Same for photos.
People will stop believing photos and video that don’t have a verifiable signature. Social networks and news organizations will automatically verify the signatures of all photos and videos they display.
Technically this is already possible today, it just needs to become mainstream and the default.
Even that isn’t possible. While you could confirm it hasn’t been modified via hashing, it can only confirm that after it was created. If you created an entirely new file there’s no way to prove it wasn’t faked and then had a signature applied.
Also, what’s to stop you from generating new hash information that is consistent with the new media if you do modify an existing one?
You’d need secure chips that can’t reveal the key, and those would be signed by a trusted authority.
Then there’d be a black market for valid chips, or maybe some tomfoolery to make a camera think it’s seeing something that’s being fed into it via a different input.
Oh yeah, forgot about cryptographic hashing for a moment there. Though that “trusted authority” is the weak point, probably only a matter of time before it is corrupted and gives keys out to powerful others.
Or same thing for any workers with access to either the central key or one of the others. And if a specific company’s key gets leaked, does that mean anything produced by their devices can no longer be trusted? If there’s an inconvenient video in existence, will the way of defeating it just be to leak the private key protecting its hashes and just say that the hackers must have gotten ahold of the key earlier than everyone else to explain how that video existed before the key was leaked publically?
Or even just have someone break in to each of their systems and steal the keys but leave evidence of it happening and use that to create reasonable doubt about whatever videos they want to call fake news.
Like maybe taking a video of a screen playing a video/a video projected on a surface?
The timestamp of the original hash being sent to a central server.
That’s the whole point of sending that hash close to when it happens.
It has always been possible to fake a video or photo, or to lie when testifying. The solution will be the same as always: presenting multiple forms of evidence, investigating people suspected of lying, and having a high standard of proof for criminal cases. AI might mean an increase in faked videos and photos, but nothing that can’t be done already.
I’m not sure it’ll change much. We already know confessions, lineups, and nearly all “forensic sciences” are unreliable at best; frequently outright false. But they’re all still used, and wrongfully ruin peoples lives.
All they need is an “expert” to testify that, “by their judgment” a video is real.
I just want to take this opportunity to state that police drug dogs have a nearly 80% false positive rate.
AI / Deepfake forensics will become a growing field.
How could any forensics be performed to discern AI from non-AI? Would it be no more than counting the fingers?
That’s one way AI-generated image leaves artifacts and clues. It could even be traces that cannot be seen by human eyes. There will be other signals and it will probably be arms race where forensics will improve, but AI will get even better, so on and so forth. Time will tell.
Isn’t there already a shit ton of laws regarding what kind of video are admissible in court? Stuff like CCTV recorder need a sealed hard drive that can be open only by acredited security guards and cops.
People have been forging photos and video as soon as they’ve been invented.
All evidence is about convincing a judge/jury it’s real. Cases almost never rely solely on photo/video evidence right now, so it won’t be some existential legal apocalypse. Prosecutors will just have to be more thorough about corroborating evidence and building a well-rounded case. Defendants will have an easier time throwing out one small part of the case.
We do the same as we did before the invention of cameras. rely on witnesses
Witness testimony was never very good, and that’s not going to change. Humans are too fallible, their testimony is too easy to attack. We don’t necessarily go back to something that worked like shit in the first place.
Fortunately fingerprints and DNA are still harder to fake. For now.
Fingerprints are a faux science as it is anyway, pretty much for the same reasons witness testimony is nearly worthless. Vsauce2 did a really good video discussing fingerprint evidence and how the specialist are very frequently wrong. Not to mention that the “all humans have unique fingerprints” idea was debunked.
DNA evidence is really the only fool-proof evidence we have.
I was talking specifically about alternative to video cameras. DNA and Finger Printi/ DNA are only relevantforr certain type of crimes that involve close contact (rape, murder, roberry) but they are not alternative to cameras. You can’t use them for crimes like hit&run, Firing projectiles, abduction , deciding who is at fault in an accident, showing police using force without reason, someone accepting bribe , …
DNA, Fingerprint are complementary tools not replacement
There is a technical solution: Have cameras sign the raw data recorded from the sensor in situ, which allows you to check if a given video was recorded on a given device. You could even add another layer where the device uploads timestamps, frame hashes and location data to a blockchain (finally a usecase!) to pin it to a certain place and time. Sure, all of this can be circumvented with enough effort, but it will make it much harder. Especially if a video can be pinned to a place, you can verify if it matches that scenery.
I don’t want a camera with a tpm, that perhaps even uploads my images to a third party. Also, if the cam has not connection that picture will just be counted as invalid? If a signing key will inevitably be leaked, authentic unauthentic Images will happen.
I think this system is flawed beyond repair.
I think this system is flawed beyond repair.
Of course, but will that stop somebody from trying to push it? ;)
I expect that if/when that time comes, we’ll see the credibility of video evidence decline.
Currently at my night job, there is video surveillance of common areas. Because of the skill and tools required in doctoring/deepfaking, there’s a pretty large window in which footage could be used should it ever need to be. But when we approach the point at which doctoring/deepfakes become pedestrian or even automated, that window slims dramatically.
Depending on the speed and ease with which it can be done, we could see legitimate arguments in favor of discounting video evidence as it then becomes less reliable than eyewitness testimony (which is already notoriously unreliable).
Maybe people will start to get out more
