The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
Going deep on deep fakes
Plus an interview with Rob Silvers on the Cyber Safety Review Board
This was a big week for AI-generated deep fakes. Sultan Meghji, who's got a new AI startup of his own, walked us through four stories that illustrate how AI will lead to more confusion about what's real and what's not. First, a fake Biden robocall urged people not to vote in the New Hampshire primary. Second, a bot purporting to offer Dean Phillips's views on the issues was penalized by OpenAI because it didn't have Phillips's consent. Third, fake nudes of Taylor Swift led to a ban on Twitter searches for her image. And, finally, podcasters used AI to resurrect George Carlin and got sued by his family for violating copyrightish law. The moral panic over AI fakery meant that all of these stories were too long on "end of the world" and too short on "we'll live through this."
Regulators of AI are not doing a much better job of maintaining perspective. Mark MacCarthy reports that New York City's AI hiring law, which has punitive disparate-impact disclosure requirements for automated hiring decision engines, seems to have persuaded NYC employers, conveniently, that none of them are using automated hiring decision enginess, so they don't have to do any disclosures. Not to be outdone, the European Court of Justice has decided that pretty much any tool to aid in decisions is an automated decision making technology subject to special (and mostly nonsensical) data protection rules.
Is AI regulation beginning to suffer from backlash? Could be. Sultan and I report on a very plausible Republican plan to attack the Biden AI executive order on the ground that its main enforcement mechanism, the Defense Production Act, simply doesn't authorize the measures the order calls for.
In other Big Tech regulation, Maury Shenk explains the EU's application of the Digital Markets Act to tech companies like Apple and Google. Apple isn't used to being treated like just another tech company, and its contemptuous response to the EU's rules for its app market could easily spur regulatory sanctions. Looking at Apple's proposed compliance with the California court ruling in the Epic case and the European Digital Market Act, Mark says it's time to think about price regulating mobile app stores.
Even handing out big checks to technology companies turns out to be harder than it first sounds. Sultan and I talk about the slow pace of payments to chip makers, and the political imperative to get the deals done before November (and probably before March).
Senator Ron Wyden, D-Ore. is still flogging NSA and the danger of government access to personal data. This time, he's on about NSA's purchases of commercial data. So far, so predictable. But he's also misrepresenting the facts by claiming flatly that NSA buys domestic metadata, ignoring NSA's clear statement that the metadata it buys is "domestic" only in the sense that it covers communications with one end inside the country. Communications with foreign countries that flow into and out of the U.S. have long been considered appropriate foreign intelligence targets, as witness the current debate over FISA section 702.
Maury and I review a Jim Dempsey's effort to construct a liability regime for insecure software. His proposal looks reasonable, but Maury reminds me that he and I produced something similar twenty years ago, that is still not even close to adoption anywhere in the U.S.
I can't help but rant about Amazon's arrogant, virtue-signaling, and customer-hating decision to drop a feature that makes it easy for Ring doorbell users to share their videos with the police. Whose data is it, anyway, Amazon? Sadly, I'm afraid we know the answer.
It looks as though there's only one place where hasty, ill-conceived tech regulation is being rolled back. China. Maury reports on China's decision to roll back video game regulations, to fire its video game regulator, and to start approving new games at a rapid clip -- though only after a regulatory crackdown had knocked more than $60 billion off the value of its industry.
We close the news roundup with a few quick hits:
- Outside of AI, VCs are closing their wallets and letting startups run out of money
- Apple launched what looks like an expensive dud – the Vision Pro
- Quantum winter may be back as quantum computing turns out to be harder than hoped
- And, speaking of winter, consumers and regulators seem to be cooling fast on self-driving cars in the wake of serious missteps by Cruise
Finally, as a listener bonus, we hear from Rob Silvers, Under Secretary for Policy at the Department of Homeland Security and Chair of the Cyber Safety Review Board (CSRB). Under Rob's leadership, DHS has proposed legislation to give the CSRB a legislative foundation. The Senate homeland security committee recently held a hearing about that idea. Rob wasn't invited, so we asked him to come on the podcast to respond to issues that the hearing raised – conflicts of interest, subpoena power, choosing the incidents to investigate, and more.
You can subscribe to The Cyberlaw Podcast using iTunes, Google Play, Spotify, Pocket Casts, or our RSS feed. As always, The Cyberlaw Podcast is open to feedback. Be sure to engage with @stewartbaker on Twitter. Send your questions, comments, and suggestions for topics or interviewees to CyberlawPodcast@gmail.com. Remember: If your suggested guest appears on the show, we will send you a highly coveted Cyberlaw Podcast mug! The views expressed in this podcast are those of the speakers and do not reflect the opinions of their institutions, clients, friends, families, or pets.
Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Comments may only be edited within 5 minutes of posting. Report abuses.
Please
to post comments
"Whose data is it, anyway, Amazon?"
Well, whose server is it on?
NYT reporters make up quotes,
AI makes up pictures.
Show me the difference.
No one should trust photos / 'film' clips any more than they trust the source. It's all bullshit these days.
One obvious difference is that the AI doesn't know it's being deceitful. **rimshot**
AI is ideal for people who already firmly believe that the world is an artificial construct made by people deliberately out to deceive them for highly complex, arcane and malevolent reasons (as opposed to the fact that mostly it's because they fired all the reporters and journalists to keep dividends high) because AI can theoretically generate any world they want, such as a world where Trump won and Taylor Swift is a CIA asset.
AI seems to be (so far, at least) for people who struggle with language, interpersonal relationships, creativity, nuance, authority, and the reality-based world.
It's a "moral panic" to fear being defamed by a doctored photograph? Not sure why photos are privileged and texts aren't if they can both communicate the same sort of defamation of an innocent person.
Geez. Type “ Taylor Swift nude” on most any search engine and you’ll get a ton of images of varying quality. Nudes of Taylor Swift is not an Elon Musk ploy.
On the other hand, Right-Wing World is awash in Taylor Swift conspiracy fever. We hear on Fox News she's a Deep-State Pentagon operative. Other wingnuts believe the Chief's victory was rigged - along with the whole Kelce business - to set-up a halftime endorsement of Biden which swings the election. There's quite the panic about that.....
Hey - sounds good to me! And if the Right's hive mind is gonna work itself into a tizzy over nonsense (as it always does), better Ms Swift than any of the old stuff - like CRT in the public schools or the national danger of transexuals.
https://www.msnbc.com/top-stories/latest/taylor-swift-fox-news-pentagon-conspiracy-theory-rcna133283
Trump has declared 'holy war' on her. What the fuck is wrong with these people?
What the fuck is wrong with these people?
I suspect that that they have to keep saying dumber and dumber stuff to keep their paychecks coming. After all, their audience is not exactly reality-based.
But this is 'fuck the election' level dumb, and not even for evil and unpopular policy goals, like abortion bans and trans persecution, it's just petty male fragility. If he goes after Beyonce, that's it, it's all over.
It’s all about entertainment. That’s the Right’s focus: Delivering a rousing consumer product from their pet media and the politician extras on the set. Most of the latter are struggling to get their name in the credits or even billing as a top performer. Be just a Ordinary Joe working on drab policy and you’re a squish or RINO by default. Those guys don’t provide any good TV viewing.
Reality? It only gets in the way. The best analogy remains pro wrestling, where spectators scream themselves hoarse over an obvious fraud. Today’s Right-wing consumer wants the pounding excitement of righteous rage and ecstasy of shrieking for their glorious heroes. The former seems inevitably purchased with victimhood, murky conspiracies, and national “emergencies” pitched to a florid hysterical level. The latter are usually crude frauds. But it all gets their wee little hearts beating wildly and that’s the only thing that matters.
As for history, it’s simple : The emergence of talk radio gave the Right-wing consumer a daily few hours of entertainment that paralleled reality. Fox News upped infotainment to 24/7. The internet then offered whole new horizons of political porn. That’s how we ended up with the Right’s Reality-TV show issues and politicians. It’s Housewives of New Jersey setting the national agenda, but the base finds it all stirring television.
I don't think so. It's about targeting anyone perceived as in their way or against them, but with no moderation or consideration, believing they can bulldoze everyone and everything into submission. They enjoy it of course, or perform their enjoyment, at least.
.
I hope that is parody. I would not wager against it, though.
On the other hand, the left believes birds aren't real.
https://www.nytimes.com/2021/12/09/technology/birds-arent-real-gen-z-misinformation.html
Yes. That's right. They don't. They really, really don't. No, it's absolutely true, they don't. You genius. You paragon. You colossus of the intellect.
On the other other hand, from your linked article:
Except that the creator of Birds Aren’t Real and the movement’s followers are in on a joke: They know that birds are, in fact, real and that their theory is made up.
The lack of qualifier before “the movement’s followers” is how one can tell from even a single sentence that it’s a friendly story. What’s their evidence for saying that’s generally or even broadly true?
Contrast also the claims since 2016 that Donald Trump was a Russian asset.
No, no, you go ahead and believe they don't believe birds are real. Stick with it, it'll justify you believing every batshit thing Trump tells you, because 'the left believes birds are fake.' Why the fuck not.
I guess if you're so cocksure about this you may not have actually read the the article itself, which quotes the founder smugly triumphing that some people have indeed taken it seriously -- including at least one news channel:
Indeed, but that fully begs the question.
Yes, Michael P, for one.
Oh, he was that news reporter? Who knew?
No, he believed *after reading the article.*
He believed that leftists believe it, as intended by Mr. Gaydos. I'm not seeing the problem.
Oh, you too? The only Swifts the left believes in is Taylor, right? Sap.
‘as intended’
I’m not sure he specifically intended idiots on the right to conclude that the left beleives birds are fake, but it’s presumably an interesting result.
I cannot imagine just how happy Peter McIndoe would be to read this exchange.
Boy, are you stupid. Amazing!!
Easily gulled.
That just shows she gets targeted by this sort of thing all the time. How much of that evil shit was on twitter/x prior to this deepfake crap?
Photorealistic nudes of famous women is a whole new wonderful world.
Screw you, suing lawyers!
Has Bone Saw Baker ever disclosed any of the clients (and not just the unsavory ones) whose commercial interests he advances as a paid mouthpiece at the Volokh Conspiracy?
The moral panic about deep fakes will be used to pass some entirely vague law about drawing pictures using AI, which will then be used to go after political dissidents.
ML, Have Conspiracies, Will Travel......
Thread to thread, the vicitimhood is strong in this one!
Nothing to see here. Publication of defamatory and fraudulent material related to the 2020 election has so far generated ~$1 billion and counting in adjudicated damages. Time for everyone to toughen up and get used to it. Everyone knows that's just the norm on the internet, and nobody takes it seriously. No more moral panics.
Note to the conspirators:
AI is going to be a great big problem. One that needs serious consideration. Stewart Baker is not up to the job. Please hunt around and find someone—probably more than one—expert enough, thoughtful enough, and non-conflicted enough to help. Treat it as an emergency. It's already an emergency for the nation.
.
Mr. Baker is an authoritarian, stale, snarky, fringe voice who publishes as an ostensibly libertarian forum (often, I sense, with a partisan and commercial ax to grind). He was selected for that position by Prof. Volokh and has been permitted to shill for partisan and pecuniary interests for years. Do you expect anyone at the Volokh Conspiracy to recognize, care about, or address the problem?
Actually, the bulk of that figure wasn't adjudicated at all.
Some of these problems don't appear to be the fault of AI. Anyone who can imitate Joe Biden's voice can pull off that robocall scam. Same with a lot of these fake photos.
To the extent that they were or weren't actually done with AI, this is true, and those fakes and scams are already mostly illegal. The George Carlin thing, for example, was not generated by an 'AI,' but even after this was admitted, the family is still going ahead with their court case. The Swift deepfakes, apart from their inherent awfulness, are something of a lesson in why sites need proper moderation, actually.
Griffin, your comment ignores a key point. Today's issue is not whether a fake is possible, if you find the right expert. Today's issue is that you don't need the expert anymore. And not only that, but the ordinary non-expert can turn out stuff many times faster than the expert can. So expect the size of the problem to multiply by, literally, orders of magnitude.
Stewart, despite being roughly my age, continues to be a smirking, immature high school kid, dismissing all those deep fakes, which not so coincidentally always seem to be concocted by Republicans/conservatives and aimed at liberals/women. One wonders how he would feel if it was done to him or to one of his dormmates.
He's a paid lobbyist for NGO Group. Yeah, the company that helps authoritative governments murder dissidents.
Don't bother appealing to his conscience.
Bone Saw Baker serves a worthwhile purpose -- he refutes anyone who might still try to claim this old, white, male, right-wing, "academic" blog is not a falsely advertised, faux libertarian clingerfest.