The White House wants to 'cryptographically verify' videos of Joe Biden so viewers don't mistake them for AI deepfakes

Depending on the implementation, there are two cryptographic functions that might be used (perhaps in conjunction):

• Cryptographic hash: An arbitrary amount of data (like a video file) is used to create a “hash”—a shorter, (effectively) unique text string. Anyone can run the file through the same function to see if it produces the same hash; if even a single bit of the file is changed, the hash will be completely different and you’ll know the data was altered.
• Public key cryptography: A pair of keys are created, one of which can only encrypt data (but can’t decrypt its own output), and the other, “public” key can only decrypt data that was encrypted by the first key. Users (like the White House) can post their public key on their website; then if a subsequent message purporting to come from that user can be decrypted using their public key, it proves it came from them.

The best way this could be handled is a green check mark near the video that you could click on it and it would give you all the meta data of the video (location, time, source, etc) with a digital signature (what would look like a random string of text) that you could click on and your browser would show you the chain of trust, where the signature came from, that it's valid, probably the manufacturer of the equipment it was recorded on, etc.

it would potentially be associated with a law that states that you must not misrepresent a “verified” UI element like a check mark etc, and whilst they could technically add a verified mark wherever they like, the law would prevent that - at least for US companies

it may work in the same way as hardware certifications - i believe that HDMI has a certification standard that cables and devices must be manufactured to certain specifications to bear the HDMI logo, and the HDMI logo is trademarked so using it without permission is illegal… it doesn’t stop cheap knock offs, but it means if you buy things in stores in most US-aligned countries that bear the HDMI mark, they’re going to work

For the average end-user, it would look like "https". You would not have to know anything about the technical background. Your browser or other media player would display a little icon showing that the media is verified by some trusted institution and you could learn more with a click.

In practice, I see some challenges. You could already go to the source via https, EG whitehouse.gov, and verify it that way. An additional benefit exists only if you can verify media that have been re-uploaded elsewhere. Now the user needs to check that the media was not just signed by someone (EG whitehouse.gov. ru), but if it was really signed by the right institution.

Adobe is actually one of the leading actors in this field, take a look at the Content Authenticity Initiative (https://contentauthenticity.org/)

Like the other person said, it’s based on cryptographic hashing and signing. Basically the standard would embed metadata into the image.

It needs some kind of handler, but we mostly have those in place. A web browser could be the handler for instance. A web browser has the green dot on the upper left, telling you a page is secure, that https is on and valid. This could work like that, the browser can verify the video and display a green or red dot in the corner, the user could just mouse over it/tap on it to see who it's verified to be from. But it's up to the user to mouse over it and check if it says whitehouse.gov or dr-evil-mwahahaha.biz

TL;DR: one day the user will see an overlay or notification that shows an image/movie is verified as from a known source. No extra software required.

Honestly, I can see this working great in future web browsers. Much like the padlock in the URL bar, we could see something on images that are verified. The image could display a padlock in the lower-left corner or something, along with the name of the source, demonstrating that it's a securely verified asset. "Normal" images would be unaffected. The big problem is how to put something on the page that cannot be faked by other means.

It's a little more complicated for software like phone apps for X or Facebook, but doable. The problem is that those products must choose to add this feature. Hopefully, losing reputation to being swamped with unverifiable media will be motivation enough to do so.

The underlying verification process is complex, but should be similar to existing technology (e.g. GPG). The key is that images and movies typically contain a "scratch pad" area in the file for miscellaneous stuff (metadata). This is where the image's author can add a cryptographic signature for the file itself. The user would never even know it's there.

Probably you'd notice a bit of extra time posting for the signature to be added, but that's about it, the responsibility for verifying the signature would fall to the owners of the social media site and in the circumstances where someone asks for a verification, basically imagine it as a libel case on fast forward, you file a claim saying "I never said that", they check signatures, they shrug and press the delete button and erase the post, crossposts, and if it's really good screencap posts and those crossposts of the thing you did not say but is still being attributed falsely to your account or person.

It basically gives absolute control of a person's own image and voice to themself, unless a piece of media is provable to have been made with that person's consent, or by that person themself, it can be wiped from the internet no trouble.

Where it comes to second party posters, news agencies and such, it'd be more complicated but more or less the same, with the added step that a news agency may be required to provide some supporting evidence that what they said is not some kind of misrepresentation or such as the offended party filing the takedown might be trying to insist for the sake of their public image.

Of course there could still be a YouTube "Stats for Nerds"-esque addin to the options tab on a given post that allows you to sign-check it against the account it's attributing something to, and a verified account system could be developed that adds a layer of signing that specifically identifies a published account, like say for prominent news reporters/politicians/cultural leaders/celebrities, that get into their own feed so you can look at them or not depending on how ya be feelin' that particular scroll session.

Rather that using a hash of the video data, you could just include within the video the timestamp of when it was originally posted, encrypted with the White House’s private key.

Apple's scrapped on-device CSAM scanning was based on perceptual hashes.

The first collision demo breaking them showed up in hours with images that looked glitched. After just a week the newest demos produced flawless images with collisions against known perceptual hash values.

In theory you could create some ML-ish compact learning algorithm and use the compressed model as a perceptual hash, but I'm not convinced this can be secure enough unless it's allowed to be large enough, as in some % of the original's file size.

As soon as VLC adds the gpg sig feature, it's over.

Yeah but all it takes is proving it doesn't have the right signature and you can make the Social Media corpo take every piece of media with that signature just for that alone.

What's even better is that you can attack entities that try to maliciously let people get away with misusing their look and fake being signed for failing to defend their IP, basically declaring you intend to take them to court to Public Domainify literally everything that makes them any money at all.

If billionaires were willing to allow disinformation as a service then they wouldn't have gone to war against news as a service to make it profitable to begin with.

It doesn't help that in a lot of cases, this is actually accepted by a shit ton of important institutions that should be better, but aren't.

Still trying to get people to sign their emails lol

There is, but only if you can implement a layer of abstraction and get them to trust that layer of abstraction.

Few laymen understand why Bitcoin is secure. They just trust that their wallet software works and because they were told by smarter people that it is secure.

Few laymen understand why TLS is secure. They just trust that their browser tells them it is secure.

Few laymen understand why biometric authentication on their phone apps is secure. They just trust that their device tells them it is secure.

Yikes! The implication is that it does not matter if a child was victimized. It's "heinous", not because of a child's suffering, but because... ?

Idk, making CP where a child is raped vs making CP where no children are involved seem on very different levels of bad to me.

Both utterly repulsive, but certainly not exactly the same.

One has a non-consenting child being abused, a child that will likely carry the scars of that for a long time, the other doesn't. One is worse than the other.

E: do the downvoters like... not care about child sexual assault/rape or something? Raping a child and taking pictures of it is very obviously worse than putting parameters into an AI image generator. Both are vile. One is worse. Saying they're equally bad is attributing zero harm to the actual assaulting children part.

I honestly do not see the value here. Barring maybe a small minority, anyone who would believe a deepfake about Biden would probably also not believe the verification and anyone who wouldn't would probably believe the administration when they said it was fake.

The value of the technology in general? Sure. I can see it having practical applications. Just not in this case.

The problem is that even if you reveal the video as fake,the feeling it reinforces on the viewer stays with them.

"Sure that was fake,but the fake that it seems believable tells you everything you need to know"

I mean, how is anyone going to crytographically verify a video? You either have an icon in the video itself or displayed near it by the site, meaning nothing, fakers just copy that in theirs. Alternatively you have to sign or make file hashes for each permutation of the video file sent out. At that point how are normal people actually going to verify? At best they're trusting the video player of whatever site they're on to be truthful when it says that it's verified.

Saying they want to do this is one thing, but as far as I'm aware, we don't have a solution that accounts for the rampant re-use of presidential videos in news and secondary reporting either.

I have a terrible feeling that this would just be wasted effort beyond basic signing of the video file uploaded on the official government website, which really doesn't solve the problem for anyone who can't or won't verify the hash on their end.

Maybe some sort of visual and audio based hash, like musicbrainz ids for songs that are independant of the file itself but instead on the sound of it. Then the government runs a server kind of like a pgp key server. Then websites could integrate functionality to verify it, but at the end of the day it still works out to a "I swear we're legit guys" stamp for anyone not techinical enough to verify independantly thenselves.

I guess your post just seemed silly when the end result of this for anyone is effectively the equivalent of your "signed by trump" image, unless the public magically gets serious about downloading and verifying everything themselves independently.

Fuck trump, but there are much better ways to shit on king cheeto than pretending the average populace is anything but average based purely on political alignment.

You have to realize that to the average user, any site serving videos seems as trustworthy as youtube. Average internet literacy is absolutely fucking abysmal.

It needs to focus on showing who published it, not the icon

That's not the point. It's that malicious actors could easily exploit that lack of knowledge to trick users into giving fake videos more credibility.

If I were a malicious actor, I'd put the words "✅ Verified cryptographically by the White House" at the bottom of my posts and you can probably understand that the people most vulnerable to misinformation would probably believe it.

Back in the day, many rulers allowed only licensed individuals to operate printing presses. It was sometimes even required that an official should read and sign off on any text before it was allowed to be printed.

Freedom of the press originally means that exactly this is not done.

As long as the source of trust is independent of the government or even big business, it would work and be trustworthy

That sounds like wishful thinking

We've had this discussion a lot in the Bitcoin space. People keep arguing it has to change so that "grandma can understand it" but I think that's unrealistic. Every technology has some inherent complexities that cannot be removed and people have to learn if they want to use it. And people will use it if the motivation is there. Wifi has some inherent complexities people have become comfortable with. People know how to look through lists of networks, find the right one, enter the passkey or go through the sign on page. Some non-technical people know enough about how Wifi should behave to know the internet connection might be out or the route might need a reboot. None of this knowledge was commonplace 20 years ago. It is now.

The knowledge required to leverage the benefits of cryptographic signatures isn't beyond the reach of most people. The general rules are pretty simple. The industry just has to decide to make the necessary investments to motivate people.

The number of 80 year olds that know what cryptography is AND know that it's a proper solution here is not large. I'd expect an 80 year old to say something like "we should only look at pictures sent by certified mail" or "You cant trust film unless it's an 8mm and the can was sealed shut!"

The President's job isn't really to be an expert on everything, the job is more about being able to hire people who are experts.

If this was coupled with a regulation requiring social media companies to do the verification and indicate that the content is verified then most people wouldn't need to do the work to verify content (because we know they won't).

It obviously wouldn't solve every problem with deepfakes, but at least it couldn't be content claiming to be from CNN or whoever. And yes someone editing content from trusted sources would make that content no longer trusted, but that's actually a good thing. You can edit videos to make someone look bad, you can slow it down to make a person look drunk, etc. This kind of content should not considered trusted either.

Someone doing a reaction video going over news content or whatever could have their stuff be considered trusted, but it would be indicated as being content from the person that produced the reaction video not as content coming from the original news source. So if you see a "news" video that has it's verified source as "xXX_FlatEarthIsReal420_69_XXx" rather than CNN, AP News, NY Times, etc, you kinda know what's up.

If a news outlet is found signing a fake video, they will be in trouble.

I see you've never heard of Fox News before.

https://en.wikipedia.org/wiki/Fox_News_controversies#Video_footage_manipulation

Not trusting unsigned videos is one thing, but will people be judging the signature or the content itself to determine if it is fake?

Why only one source should be trusted is a salient point. If we are talking trust: it feels entirely plausible that an entity could use its trust (or power) to manufacture a signature.

And for some it is all too relevant that an entity like the White House, (or the gambit of others, past or present), have certainly presented false informstion as true to do things like invade countries.

Trust is a much more flexible concept that is willing to be bent. And so cryptographic verification really has to demonstrate how and why something is fake to the general public. Otherwise it is just a big 'trust me bro.'

Great points and I agree. I also think the signature needs to be built into the stream in a continuous fashion so that snippets can still be authenticated.

I don't think that's practical or particularly desirable.

Today, when you buy something, EG a phone, the brand guarantees the quality of the product, and the seller guarantees the logistics chain (that it's unused, not stolen, not faked, not damaged in transport, ...). The typical buyer does not care about the parts used, the assembly factory, etc.

When a news source publishes media, they vouch for it. That's what they are paid for (as it were). If the final viewer is expected to check the chain, they are asked to do the job of skilled professionals for free. Do-your-own-research rarely works out, even for well-educated people. Besides, in important cases, the whole chain will not be public to protect sources.

I've thought about this too but I'm not sure this would work. First you could hack the firmware of a cryptographically signed camera. I already read something about a camera like this that was hacked and the private key leaked. You could have an individual key for each camera and then revoke it maybe.

But you could also photograph a monitor or something like that, like a specifically altered camera lens.

Ultimately you'd probably need something like quantum entangled photon encoding to prove that the photons captured by the sensor were real photons and not fake photons. Like capturing a light field or capturing a spectrum of photons. Not sure if that is even remotely possible but it sounds cool haha.

Look up transparency logs for that last part, it's already used for TLS certificates

The point is that if someone catches the President shagging kids, of course that footage won't be authenticated by the WH. We need a tool so that a genuine piece of footage of the Pres shagging kids would be authenticated, but a deepfake of the same would not. The WH is not a good arbiter since they are not independent.

Then this exercise is a waste of time. All the hard hitting journalism which presses the President and elicits a negative response will be unsigned, and will be distributed across social media as it is today: without authentication. All the videos for which the White House is concerned about authenticity will continue to circulate without any cause for contention.

That's fine?

Signatures aren't meant to prove authenticity. They're proving the source which you can use to weigh the authenticity.

I think the confusion comes from the fact that cryptographic signatures are mostly used in situations where proving the source is equivalent to proving authenticity. Proving a text message is from me proves the authenticity as there's no such thing as doctoring my own text message. There's more nuance when you're using signatures to prove a source which may or may not be providing trustworthy data. But there is value in at least knowing who provided the data.

Might just as well show a dog a card trick.

Should probably start out with the colour mixing one. That was very helpfull for me to figure out public key cryptography. The difficulty comes in when they feel like you are treating them like toddlers so they start behaving more like toddlers. (Which they are 99% if the time)

I can totally see this being a thing and I kinda wish it would just because I love old people trying to seem like they know tech when they don't but in the context of still helpful tech stuff.

Thank you, lol. This is what people end up with when they think of the first solution that comes to mind. Often just something that makes life harder for everyone EXCEPT bad actors. This just creates hoops for people following the rules to jump though while giving the impression the problem was solved, when it's not.

You must be severely misunderstanding the idea. The idea is not to encrypt it in a way that it's only unlockable by a secret and hidden key, like DRM or cable TV does, but to do the the reverse - to encrypt it with a key that is unlockable by publicly available and widely shared key, where successful decryption acts as a proof of content authenticity. If you don't care about authenticity, nothing is stopping you from spreading the decrypted version, so It shouldn't affect consumers one bit. And I wouldn't describe "Get a bunch of cameras, rip the sensors out, carefully and repeatedly strip the top layers off and scan using electron microscope until you get to the encryption circuit, repeat enough times to collect enough scans undamaged by the stripping process to then manually piece them together and trace out the entire circuit, then spend a few weeks debugging it in a simulator to work out the encryption key" as "trivial"

Oh, they've actually been developing that! Thanks for the link, I was totally unaware of C2PA thing. Looks like the ball has been very slowly rolling ever since 2019, but now that the Google is on board (they joined just a couple days ago), it might fairly soon be visible/usable by ordinary users.

Mark my words, though, I'll bet $100 that everyone's going to screw it up miserably on their first couple of generations. Camera manufacturers are going to cheap out on electronics, allowing for data substitution somewhere in the pipeline. Every piece of editing software is going to be cracked at least a few times, allowing for fake edits. And production companies will most definitely leak their signing keys. Maybe even Intel/AMD could screw up again big time. But, maybe in a decade or two, given the pace, we'll get a stable and secure enough solution to become the default, like SSL currently is.

Only RSA uses a function equivalent to encryption when producing signatures, and only when used in one specific scheme. Every other algorithm has a unique signing function.

Thanks a lot, that helped me understand. Seems like a good idea

Digital signature. A watermark may be useful so that an unauthorized user can’t easily hide their source without noticeably defacing the photo, but it doesn’t prevent anyone from modifying it

A digital signature is a somewhat similar idea except that signature verification fails if there are any changes. This is tough to do with a photograph, where some applications may be blindly re-encoding or modifying the resolution so those may need to be fixed.

You could argue this is a good use case for blockchain, certainly much better than those stupid monkey images. When John Stewart parodies a politician, there should be a verifiable chain of evidence from the White House release to the news bureau to his studio, before they alter the lighting to highlight orange skin tone for yucks.

I agree that it’s a good idea. But the people most swayed by deepfakes of Biden are definitely the least concerned with whether their bogeyman, the “deep state” has verified them

I was being glib, but as NFTs are (typically) images signed by a blockchain, it meets the criteria of "cryptographically signed image" in a way.

In reality, you are correct.

Transparency logs like that are helpful to show when media was first seen / published

Putting it on the blockchain ensures you can always go back and say "see, at this date/time, this key verified this file/hash".. If you know the key of the uploader (the white house), you can verify it was signed by that key. Guatemala used a similar scheme to verify votes in elections using Bitcoin. Could the precinct lie and put in the wrong vote count? Of course! But what it prevented was somebody saying "well actually the precinct reported a different number" since anybody could verify that on chain they didn't. It also prevented the precinct themselves from changing the number in the future if they were put under some kind of pressure.

Guys, it doesn't need to be on a block chain. Asymmetric key cryptography is enought to verify authenticity.

There is that. It's kind of like the people that are proud to vote for Biden again too. The whole thing is a fucking joke.