This conviction has nothing to do with uploading AI generated music. The illegal part was using tens-of-thousands of bot accounts to listen to them (and the ads) generating fraudulent revenue for himself.
This is just the same old view botting (aka click fraud) that has been going on for decades. The AI generated music aspect is irrelevant.
Click fraud is cut and dry wire fraud: Using electronic communications to deceptively steal money or property.
> Whoever, having devised or intending to devise any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises, transmits or causes to be transmitted by means of wire, radio, or television communication in interstate or foreign commerce, any writings, signs, signals, pictures, or sounds for the purpose of executing such scheme or artifice
This kind of fraud would be impossible if revenue from each subscriber is distributed just to the artists they listened to. Bots listening to thousands of songs would not make a difference in this model. And I would be much happier if my money went to struggling artists I like and support, rather than to the global top 10, of whom I never played a single song.
I don't have data, but my gut feeling is that it would make a significant difference to niche artists with small but loyal listeners.
>And I would be much happier if my money went to struggling artists I like and support, rather than to the global top 10, of whom I never played a single song.
I'm not a fan of anything that happened here, but at the same time I have some concerns about the ability of large corporations to convert TOS violations into federal crimes. Streaming media vendors seem to get all kinds of special attention when it comes time to prosecute other humans.
He essentially defrauded $8M from media streaming companies, if the ToS violation is the easiest path for the companies to have him quite rightfully convicted of what is indisputably criminal behaviour then I have zero qualms about that.
Does this behaviour open the door for ToS being abused? I have no legal expertise, but I would expect in cases like this that everyone would rationally come to the conclusion that the defendant's behaviour was wrong and unethical and the ToS just made it easy for the plaintiffs to point out to the court that they do in fact explicitly forbid such activities, making it an open and shut case.
From headlines I've seen of around ToS enforcement over the decades, courts don't seem to just view them the same as a physically signed legal agreement and will not enforce outrageous clauses in them.
The steaming services won't have lost any money on this.
It's the advertisers who paid for ads to get played to the bot accounts, and (depending on how the advertising deals were structured) other artists with legitimate listeners might have received smaller revenue cuts.
> rightfully convicted of what is indisputably criminal behaviour
Consider the opposite view: if pretending to be a human is "criminal behavior" there are about 8 billion criminals walking around on this planet.. and in this case our current legal system appears to be hijacked for the protection of utterly nonsensical, hopelessly broken, ancient business models from a rent-seeking, anti-consumer, creator-exploiting, trillion-dollar corporate mafia, which would like nothing better than to track, spy, and force-feed their audience at every turn.
> At certain points, SMITH had as many as 10,000 active Bot Accounts on
the Streaming Platforms
> Later, SMITH attempted to sell his fraudulent streaming scheme as a service, in which other musicians would pay him for streams he would fraudulently generate or share royalties with him in exchange for fraudulent streams of their
music
> In or about 2018, SMITH began working with the Chief Executive Officer of an AI music company ("CC-3") and a music promoter ("CC-4") to create hundreds of thousands of songs using artificial intelligence that SMITH could then fraudulently stream.
If I have three email addresses petethecoolone@gmail.com, joemama69@gmail.com, and michael.j.smith@gmail.com, are those “fake” as well, then? An email address doesn’t have to reflect your real name.
How about when I use iCloud Hide My Email to generate a unique email address when I create a new account somewhere? Is that a “fake” email address as well?
Or do they mean hacked email accounts that belonged to someone else? But then calling them “fake” email addresses still seems weird wording.
A "fake" e-mail address is one that is not clearly tied to your real identity, citizen. There is no reason to use such non-identifying addresses unless you have something to hide. Comply, for the children.
Maybe a "fake" email address in their terms is an impossibly invalid one, unowned one that cannot be verified, or a disposable verifiable one? I'm not sure.
Fun fact: Gmail address prefixes can optionally intercalate a period between any letters. All accounts though must be remain unique after normalizing case and removing all periods.
a.bc@ = ab.c@ = a.b.c@ = abc@, but only one of these can be registered.
Which "spec"? Many specifications exist and not every MTA, mailbox, nor email client understands the latest and every implementation is guaranteed to have quirks, but here are the most common: RFCs 822, 2822, 4952, 5336, 6530. There is no the spec. Subaddressing isn't standardized, but there was a half-hearted RFC on that with too many extra features. How email user parts are interpreted is up to the MTA, mailbox, or email client. Plus subaddressing informally as a hack has existed since at least 2005 according to my rusty memory. It's, therefore, a de facto standard. Use it where possible to track down leaks of personal information, but it's likely to bite with some gotchas. I used to maintain a @{{name}}.name domain and email server where all emails went to a single catchall account without setup or subaddressing needed to disambiguate which source email correspondence was from.
> Dots are somewhat common as optional these days but not universal.
- AT&T “unlimited” mobile plans
- Purdue Pharma's OxyContin push
- Juul marketing vaping products as a "safer alternative" to smoking
- Facebook's sale of user data to Cambridge Analytica
- Wells Fargo opening fake accounts for people
- ...
How many Sacklers are in jail for what they did to people? None. Purdue pleaded guilty, but no Sackler family member went to jail. The settlement totals about $7.4 billion, with roughly $6.5–$7 billion coming from the Sacklers and about $900 million from Purdue. Earlier estimates put the family's wealth around $11 billion, so they remain enormously wealthy. Hundreds of thousands have died in the opioid crisis, ruined families got no real justice, and no Sackler went to prison... great punishment.
Flooding stream services with slop and autoplaying it through a bot farm is obviously bad behavior, but is it illegal, punishable with jail time (5 years mentioned)?
I see no victims other than large streaming services who failed to account for a changing reality.
I’m getting ‘because of torrenting metallica won’t be able to afford its third private jet’ vibes from this
He indisputably defrauded $8M from these companies by "tricking" them into giving it to him.
Whereas with pirating by downloading a song, the "damage" is completely hypothetical, it's not like the downloader got actual money from doing the download and it's far from certain they would have paid the normal fee if the piracy option was not available. It's unproveable that the publisher actually lost any money from the activity.
However, hosting a website offering piracy through listing of e.g. torrents where they make significant money from ad-revenue is clearly a case of you profiting off the work of others, but it's probably still a bit grey in terms of linking the harm to the rights holder.
What's an open and closed case though is any subscription service where the website charges users in some form which grants them access to media they don't have a license to distribute and to which they don't compensate the rights holder.
There's intent, deception, and damages so it's definitely fraud. This isn't a mundane matter of creatively using someone's API in a way they don't like. He came up with a scheme to extract money from them. The ToS is the contract governing payments in this case (IIUC).
It's the difference between violating a no skateboarding sign in front of a shopping mall versus a no trespassing sign at a military base. They're both "just signs", right?
I guess forging documents and selling you a house which I don't own shouldn't be an actual crime either? The patterns of behavior in both cases are functionally indistinguishable.
This conviction has nothing to do with uploading AI generated music. The illegal part was using tens-of-thousands of bot accounts to listen to them (and the ads) generating fraudulent revenue for himself.
This is just the same old view botting (aka click fraud) that has been going on for decades. The AI generated music aspect is irrelevant.
Click fraud is cut and dry wire fraud: Using electronic communications to deceptively steal money or property.
"I am using AI agents to enjoy music while I sleep, so I don't have to".
This guy is a visionary, the judges just don't get it.
He is economic growth incarnated.
Should have used OpenClaw and they’d given him investment instead
Or perhaps OpenAI/AI startups would've even hired the guy too
In click fraud, it's the platform company defrauding the ads purchasers, by claiming the clicks are legit.
All this guy did was break TOS, and find a broken business model. No fraud. Should have found a lawyer.
> Whoever, having devised or intending to devise any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises, transmits or causes to be transmitted by means of wire, radio, or television communication in interstate or foreign commerce, any writings, signs, signals, pictures, or sounds for the purpose of executing such scheme or artifice
It’s wire fraud
What were the false or fradulent pretenses, representations, or promises?
A false representation of user interest
How is it stealing when they sent you the money?
How is fraud stealing?
This kind of fraud would be impossible if revenue from each subscriber is distributed just to the artists they listened to. Bots listening to thousands of songs would not make a difference in this model. And I would be much happier if my money went to struggling artists I like and support, rather than to the global top 10, of whom I never played a single song.
I don't have data, but my gut feeling is that it would make a significant difference to niche artists with small but loyal listeners.
Which is why it is not supported by the top 5 music distribution companies which have contracts with Spotify
>And I would be much happier if my money went to struggling artists I like and support, rather than to the global top 10, of whom I never played a single song.
Why don't we have a platform like that?
> Bots listening to thousands of songs would not make a difference in this model.
The ad revenue from the bots would be distributed. The same problem happens on Youtube.
Yes this is the solution to this problem. This is what SoundCloud is doing AFAIK.
I'm not a fan of anything that happened here, but at the same time I have some concerns about the ability of large corporations to convert TOS violations into federal crimes. Streaming media vendors seem to get all kinds of special attention when it comes time to prosecute other humans.
He essentially defrauded $8M from media streaming companies, if the ToS violation is the easiest path for the companies to have him quite rightfully convicted of what is indisputably criminal behaviour then I have zero qualms about that.
Does this behaviour open the door for ToS being abused? I have no legal expertise, but I would expect in cases like this that everyone would rationally come to the conclusion that the defendant's behaviour was wrong and unethical and the ToS just made it easy for the plaintiffs to point out to the court that they do in fact explicitly forbid such activities, making it an open and shut case.
From headlines I've seen of around ToS enforcement over the decades, courts don't seem to just view them the same as a physically signed legal agreement and will not enforce outrageous clauses in them.
The steaming services won't have lost any money on this.
It's the advertisers who paid for ads to get played to the bot accounts, and (depending on how the advertising deals were structured) other artists with legitimate listeners might have received smaller revenue cuts.
Exactly and most of the advertisers are attention monopolies anyways and only using ads to reinforce their monopolies.
> rightfully convicted of what is indisputably criminal behaviour
Consider the opposite view: if pretending to be a human is "criminal behavior" there are about 8 billion criminals walking around on this planet.. and in this case our current legal system appears to be hijacked for the protection of utterly nonsensical, hopelessly broken, ancient business models from a rent-seeking, anti-consumer, creator-exploiting, trillion-dollar corporate mafia, which would like nothing better than to track, spy, and force-feed their audience at every turn.
The Indictment has an interesting section on the detail of how he did it (end of page 5): https://www.justice.gov/usao-sdny/media/1366241/dl
> At certain points, SMITH had as many as 10,000 active Bot Accounts on the Streaming Platforms
> Later, SMITH attempted to sell his fraudulent streaming scheme as a service, in which other musicians would pay him for streams he would fraudulently generate or share royalties with him in exchange for fraudulent streams of their music
> In or about 2018, SMITH began working with the Chief Executive Officer of an AI music company ("CC-3") and a music promoter ("CC-4") to create hundreds of thousands of songs using artificial intelligence that SMITH could then fraudulently stream.
> using fake email addresses
What exactly is a “fake” email address here?
If I have three email addresses petethecoolone@gmail.com, joemama69@gmail.com, and michael.j.smith@gmail.com, are those “fake” as well, then? An email address doesn’t have to reflect your real name.
How about when I use iCloud Hide My Email to generate a unique email address when I create a new account somewhere? Is that a “fake” email address as well?
Or do they mean hacked email accounts that belonged to someone else? But then calling them “fake” email addresses still seems weird wording.
A "fake" e-mail address is one that is not clearly tied to your real identity, citizen. There is no reason to use such non-identifying addresses unless you have something to hide. Comply, for the children.
Maybe a "fake" email address in their terms is an impossibly invalid one, unowned one that cannot be verified, or a disposable verifiable one? I'm not sure.
Fun fact: Gmail address prefixes can optionally intercalate a period between any letters. All accounts though must be remain unique after normalizing case and removing all periods.
a.bc@ = ab.c@ = a.b.c@ = abc@, but only one of these can be registered.
That is all email addresses, it’s part of the spec. Aside from broken implementations of course.
Periods are optional.
Edit: Woops yeah, meant pluses. Dots are somewhat common as optional these days but not universal.
False, this is only with Gmail. And not even true for Gmail workspaces.
Which "spec"? Many specifications exist and not every MTA, mailbox, nor email client understands the latest and every implementation is guaranteed to have quirks, but here are the most common: RFCs 822, 2822, 4952, 5336, 6530. There is no the spec. Subaddressing isn't standardized, but there was a half-hearted RFC on that with too many extra features. How email user parts are interpreted is up to the MTA, mailbox, or email client. Plus subaddressing informally as a hack has existed since at least 2005 according to my rusty memory. It's, therefore, a de facto standard. Use it where possible to track down leaks of personal information, but it's likely to bite with some gotchas. I used to maintain a @{{name}}.name domain and email server where all emails went to a single catchall account without setup or subaddressing needed to disambiguate which source email correspondence was from.
> Dots are somewhat common as optional these days but not universal.
This handwaving tautology is meaningless.
He's talking about aliases, not independent email addresses.
Are you sure you mean dots and not pluses?
So let me get this straight,
Big players defraud the common people -> no prosecution
Common man defrauds the big players -> prosecution
Could you provide an example of the former?
- AT&T “unlimited” mobile plans - Purdue Pharma's OxyContin push - Juul marketing vaping products as a "safer alternative" to smoking - Facebook's sale of user data to Cambridge Analytica - Wells Fargo opening fake accounts for people - ...
All of your examples are downright terrible, but perhaps that's to be expected with these sorts of arguments.
> AT&T “unlimited” mobile plans
https://www.att.com/plans/unlimited-data-plans/
There's a huge bolded disclaimer literally in the middle of the page which says "AT&T may temporarily slow data speeds if the network is busy."
> Purdue Pharma's OxyContin push
Forced to file for chapter 11
> Juul marketing vaping products as a "safer alternative" to smoking
It is strictly true that the vaping products sold by Juul are vastly safer than smoking.
> Facebook's sale of user data to Cambridge Analytica
Not fraud.
> Wells Fargo opening fake accounts for people
https://www.justice.gov/archives/opa/pr/wells-fargo-agrees-p...
Huge penalties far exceeding their gains from this scheme conducted exclusively by low-level employees.
> Forced to file for chapter 11
How many Sacklers are in jail for what they did to people? None. Purdue pleaded guilty, but no Sackler family member went to jail. The settlement totals about $7.4 billion, with roughly $6.5–$7 billion coming from the Sacklers and about $900 million from Purdue. Earlier estimates put the family's wealth around $11 billion, so they remain enormously wealthy. Hundreds of thousands have died in the opioid crisis, ruined families got no real justice, and no Sackler went to prison... great punishment.
The big example is probably AI training not being counted as piracy.
> The streaming service Deezer said earlier this year it is receiving more than 60,000 fully AI-generated tracks daily…
Wow!
Would people botting in video games get a similar sentence?
Flooding stream services with slop and autoplaying it through a bot farm is obviously bad behavior, but is it illegal, punishable with jail time (5 years mentioned)?
I see no victims other than large streaming services who failed to account for a changing reality.
I’m getting ‘because of torrenting metallica won’t be able to afford its third private jet’ vibes from this
He indisputably defrauded $8M from these companies by "tricking" them into giving it to him.
Whereas with pirating by downloading a song, the "damage" is completely hypothetical, it's not like the downloader got actual money from doing the download and it's far from certain they would have paid the normal fee if the piracy option was not available. It's unproveable that the publisher actually lost any money from the activity.
However, hosting a website offering piracy through listing of e.g. torrents where they make significant money from ad-revenue is clearly a case of you profiting off the work of others, but it's probably still a bit grey in terms of linking the harm to the rights holder.
What's an open and closed case though is any subscription service where the website charges users in some form which grants them access to media they don't have a license to distribute and to which they don't compensate the rights holder.
Nah, that's insane. "You ticked our endolpoint KPI shit in a way we don't like" = actual crime? Gtfo
There's intent, deception, and damages so it's definitely fraud. This isn't a mundane matter of creatively using someone's API in a way they don't like. He came up with a scheme to extract money from them. The ToS is the contract governing payments in this case (IIUC).
It's the difference between violating a no skateboarding sign in front of a shopping mall versus a no trespassing sign at a military base. They're both "just signs", right?
I guess forging documents and selling you a house which I don't own shouldn't be an actual crime either? The patterns of behavior in both cases are functionally indistinguishable.
> metallica won’t be able to afford its third private jet
Napster Bad: https://m.youtube.com/watch?v=fS6udST6lbE (a classic)
Do you think fraud should be legal if it's only targeting large entities like Spotify? Would that make the world a better place?
yes it’s fraud since you get money per listens
I’m seeing the future and it’s some sort of AI gray goo all over everything.
https://en.wikipedia.org/wiki/Gray_goo
Explains the Fermi paradox
Soooo basically the internet as of today?
[dead]