Here is the complete panel lineup:

Standard Fare: Standard Technical Measures in the DMCA and Beyond

Section 512(i) of the Digital Millennium Copyright Act says, in part, that an online service has to accommodate and not interfere with standard technical measures for identifying or protecting copyrighted works in order to qualify for the DMCA safe harbors. Recent DMCA reform efforts have focused attention on this heretofore obscure portion of the statute, and they are the focus of the recently proposed Tillis-Leahy SMART Copyright Act of 2022. Just what are these “standard technical measures” and how do they relate to the DMCA, the proposed legislation, and the future of online services’ copyright obligations? Our panel will demystify this murky and often contentious area.

Thank You for Your Service: The Future of the Server Test

Fifteen years ago, Perfect 10 v. Amazon.com established the Server Test for content embedded in web pages: the publisher of a web page can be liable for infringement of the embedded content if it also hosts the content on its own servers, but not if it merely links to content hosted elsewhere. Decisions in the Southern District of New York since then, including Goldman v. Breitbart, Nicklen v. Sinclair Broadcast Group, and most recently McGucken v. Newsweek, have gone the other way. Is the Server Test any more or less relevant with today’s technology, as opposed to the technology of 2007? And is it headed for the Supreme Court … or for oblivion? Our panel will discuss.

Inclusive Access: Easing the Textbook Burden

Inclusive Access, a/k/a Equitable Access, is a new model for distributing text materials to college students digitally: every student pays the same flat fee per semester for their text materials, and the university apportions royalties to the publishers of those materials. With this scheme, students no longer need to choose their majors based on whether they can afford the textbooks. Yet some say that Inclusive Access doesn’t go far enough, that only open and free educational resources will truly improve inclusion and equity in higher education. In this session we’ll look at the progress of Inclusive Access, whether it’s the best way to achieve equity for educational materials, how publishers are participating, and the implications for copyright.

Lessons from Across the Pond: The UK Parliament’s Study on Music Streaming Economics

Last year, the UK Parliament published the results of a major study on the economics of music streaming, with many recommendations for changes in copyright laws. In this session, we’ll review the study and its conclusions and recommendations, and we’ll discuss how they might apply in the United States.

Here are the panel sessions. If you’re interested in moderating or speaking on one of these panels, please send me an email indicating your name, affiliation, panel of interest, and a brief statement of your perspective on the topic.

Thank You for Your Service: The Future of the Server Test

Fifteen years ago, Perfect 10 v. Amazon.com established the Server Test for content embedded in web pages: the publisher of a web page can be liable for infringement of the embedded content if it also hosts the content on its own servers, but not if it merely links to content hosted elsewhere. Decisions in the Southern District of New York since then, including Goldman v. Breitbart, Nicklen v. Sinclair Broadcast Group, and most recently McGucken v. Newsweek, have gone the other way. Is the Server Test any more or less relevant with today’s technology, as opposed to the technology of 2007? And is it headed for the Supreme Court … or for oblivion? Our panel will discuss.

Inclusive Access: Easing the Textbook Burden

Inclusive Access, a/k/a Equitable Access, is a new model for distributing text materials to college students digitally: every student pays the same flat fee per semester for their text materials, and the university apportions royalties to the publishers of those materials. With this scheme, students no longer need to choose their majors based on whether they can afford the textbooks. Yet some say that Inclusive Access doesn’t go far enough, that only open and free educational resources will truly improve inclusion and equity in higher education. In this session we’ll look at the progress of Inclusive Access, whether it’s the best way to achieve equity for educational materials, how publishers are participating, and the implications for copyright.

Lessons from Across the Pond: The UK Parliament’s Study on Music Streaming Economics

Last year, the UK Parliament published the results of a major study on the economics of music streaming, with many recommendations for changes in copyright laws. In this session, we’ll review the study and its conclusions and recommendations, and we’ll discuss how they might apply in the United States.

As usual, I’m producing the conference in partnership with my good friends at the Copyright Society (whose trustee board I am proud to have joined), and the Hansen IP Institute at Fordham. We’ll be livestreaming the event if you can’t make it to NYC in person.

I’m putting the agenda together. The format will be similar to our last in-person event in 2020: a keynote followed by two parallel tracks of three sessions. Please send me your suggestions for speakers and panels! I expect to post the agenda and registration information within the next couple of weeks.

We will have two one-hour sessions on each of the two days. Click on any of them to see the session description:

• Is There a Filter in Your Future? The Paths of Platform Liability in the U.S. and Europe
• Leaving the Garden: Licensing Comes to Podcasting
• Synch and Swim: Music Synch Licensing and the Online Video Explosion
• NFTs and Inclusiveness in the Art Market

Registration info will be available shortly. Meanwhile, I’m accepting proposals to speak on any of these panels now. Please send your proposal with the following information:

• The basics: proposed speaker’s name, position, organization, etc.
• A bio or CV, or a link to one (LinkedIn profile or company web page is fine)
• Session name
• Brief description of speaker’s perspective on or experience with the topic
• Contact information of proposed speaker

I will accept proposals on a rolling basis, but we intend to publish the full slate of speakers by the middle of August. We value diversity in our speaker roster. If you are submitting a proposal on behalf of someone else, please note that if the proposal is accepted, the speaker must confirm personally before they will be put on the program.

The popular narrative about the music industry, expressed in books and elsewhere, is that it was humming along fine until the Internet and Napster came along and disrupted it. That’s a cramped view of an industry that started over a century ago. Our book puts the Internet/Napster narrative in its proper context as just one of many technology-driven disruptions, which were just as dramatic in their own ways.

Here’s just one example. It’s common knowledge in the industry that file-sharing in the 1990s contributed to a nosedive in industry revenue from its peak in 1999 to less than half 15 years later. But that’s not the first time industry revenue dropped dramatically due to technological disruption. In the late 1940s, record companies created new technologies for producing records that had several times the capacity of and were more durable than 78rpm shellac discs: microgroove recording on vinyl. But while Columbia’s version of the technology kept the same form factor as 78s while increasing capacity fivefold, RCA’s version kept the 3-4 minute per side capacity of 78s while making the discs one-third the size. This led to the first “format war”—between 12-inch LPs and 7-inch 45s. Consumers were confused and turned away from recorded music. As a result, industry revenues dropped by nearly 25% in only two years, during a period of postwar economic expansion. The industry didn’t recover and return to growth until the mid-1950s, when rock & roll appeared and created a mass market for 45s that was distinct from the market for LPs, which focused on classical music and Broadway soundtracks.

Our book is organized into chapters on each of the twelve formats for distributing recorded music throughout history: player piano rolls, phonograph records, radio, vinyl, tape (8-tracks and cassettes), television, CDs, digital downloads, streaming, streaming video, and into the future with AI and voice response technology. We examine developments in each one according to a framework (created by Howie) called the 6Cs: Cutting-edge technology, Creators, Channels, Consumers, Cash, and Copyright. Each chapter uses the 6C framework to examine one of the formats and tells stories about how the new technologies disrupted the industry.

Below is an edited excerpt from the Copyright section of our chapter on radio. This was occasioned by a talk I heard earlier this week at the Copyright Society of the USA’s Annual Meeting by Shira Perlmutter, the head of the U.S. Copyright Office. She reiterated the Copyright Office’s longstanding support for enacting performance rights in sound recordings, which would require AM/FM radio stations to pay royalties to record labels and artists. America is one of the only countries in the world that doesn’t pay royalties for radio airplay. Back in January, the Copyright Office wrote a joint letter with the U.S. Patent and Trademark Office (Perlmutter’s former home before moving to the Copyright Office last year) to Congress in support of legislative reform.

Radio achieved its position of power and influence in the music industry because it was by far the most efficient way to promote recorded music for a long time. Yet its ability to do that has rested largely on a couple of quirks in U.S. copyright law.

It’s important to understand that each music track normally has two copyrights: one for the musical composition, typically owned by songwriters and music publishers; the other for the sound recording of the composition, typically owned by labels or artists. Copyright law says that copyright owners have certain exclusive rights to their creations, but copyrights in sound recordings are more limited than copyrights in musical compositions. Specifically, songwriters and music publishers have exclusive rights to public performances of their compositions, but recording artists and labels don’t have exclusive rights to public performances of their records. Radio broadcasts are considered to be public performances.

The lack of a performance right in sound recordings means that radio stations have never had to license records to play them on air. They have only had to license the musical compositions.

Arguments over whether sound recording copyrights should include performance rights date back to the 1920s. For example, the president of the musicians’ union, the American Federation of Musicians, had this to say at a 1961 congressional hearing: “[I]t is a shocking crime that people like [Philadelphia Orchestra conductor] Mr. Leopold Stokowski or Leonard Bernstein, or Louis Armstrong, or whoever the artist may be, are denied the right to receive additional fees, when money is made with his product. All you have to do is put a radio set into this room today and you can listen for hours and hours to canned music here, records received free by the broadcaster, if you please, while the men who made them are sitting home trying to figure out how to pay for their children’s education.”[i] When Congress finally enacted copyrights in sound recordings with the Sound Recording Act of 1971, it didn’t include a public performance right, despite recommendations from the U.S. Copyright Office that it be included.

As far as musical compositions are concerned, court decisions in 1940 established that radio stations could obtain records and play them on the air as many times as they wanted without having to seek permission from music publishers. In legal terms, radio stations have a compulsory license to the musical compositions performed on the records they play. Stations do have to pay royalties on musical compositions, but that is simple to do through PROs (performance rights organizations), which license public performance rights to musical compositions on behalf of large numbers of songwriters and music publishers.

ASCAP, the first PRO, formed before the start of radio. Next was BMI (Broadcast Music Inc.), which focused primarily on licensing music for radio. BMI was formed in 1939 by the National Association of Broadcasters (NAB), the radio industry’s trade association, to inject price competition as ASCAP kept raising its royalty rates. ASCAP and BMI received antitrust consent decrees from the Department of Justice in 1941; these are essentially government permission slips to operate a duopoly. (There was a third PRO, SESAC, which focused on gospel music and European composers at that time; its market share at that time was negligible compared to ASCAP and BMI.)

Today, ASCAP and BMI cover most of the market with roughly equal sized catalogs, but there are other smaller PROs that also collect royalties from radio.  SESAC still exists and has broadened its catalog to include compositions by the likes of Bob Dylan, Neil Diamond, and the rock band Rush. The other PRO that licenses broadcast radio is Global Music Rights (GMR).[1]  Veteran artist manager Irving Azoff founded GMR in 2013 by convincing a group of big-name songwriters that they could get higher royalties through a separate PRO than they were getting through ASCAP or BMI. GMR’s current catalog includes only 90 songwriters (compared to over a million for BMI), but they include Bruce Springsteen, Pharrell Williams, Prince, Don Henley and Glenn Frey (the Eagles), Pete Townshend, John Lennon, George Harrison, Bruno Mars, Smokey Robinson, and Ira Gershwin.

PROs offer blanket licenses, meaning that a radio station just has to pay each PRO a license fee and it gets a license to all of the compositions in the PRO’s repertoire. The PRO then figures out whom to pay how much for the airplay.

In other words, radio stations—unlike today’s streaming music services—have had an easy path to licensing music, enabling them to play whatever music they want whenever they want without advance permission.

Congress had reopened the question of performance rights in sound recordings in 1978, just two years after a major copyright reform bill had been signed into law, but nothing came of it,[ii] and the matter was dropped for a while. But the advent of digital radio in the 1990s brought the subject back yet again. Although labels had long viewed radio airplay as promotional, they weren’t inclined to view the new digital radio formats that way. On the contrary: they were afraid that digital audio technology, with its ability to make perfect copies, would lead to services that cannibalized record sales.

So the labels got the Copyright Office to open inquiries on the effects of digital audio transmission on music piracy in 1991, which led to Congress passing the Digital Performance Right in Sound Recordings Act (DPRA) in 1995. The DPRA established the performance right for sound recordings, but only for digital radio; it left intact the lack of a performance right in sound recordings for traditional AM and FM. We’ll see why this happened later in this section. The DPRA established a procedure for setting statutory royalties (royalties set by law, not by negotiation among parties) for digital radio services, processes that the U.S. Copyright Office runs every 4-5 years and which resemble litigations in court.

The DPRA also established certain rules for AM/FM terrestrial broadcast stations that simulcast online, to help prevent people from using them to make digital copies of music. (Currently, roughly three-quarters of FCC-licensed radio stations simulcast online.) These stations weren’t allowed, for example, to play more than four songs by the same artist in a three-hour period, or to pre-announce any specific songs.[2] The DPRA also established that interactive streaming services—which wouldn’t be in existence until several years in the future—would need to negotiate licenses with record labels and wouldn’t get statutory royalty rates.

The result of the DPRA was that all types of digital radio services—Internet pure-play, AM/FM simulcast, satellite, the music channels on cable TV services, and Muzak—got compulsory licenses to play any sound recordings they wanted as long as they pay the statutory royalties. Separate royalty rates were eventually set up for the various types of digital radio.

However, as Internet radio services got more customizable, rightsholders began to argue that they had become tantamount to interactive services. In 2001, a group of record labels sued LAUNCH Media, the company that operated LAUNCHcast (which was offered through Yahoo!). LAUNCHcast offered features for users to customize their music feeds, such as by rating music and following other users; but it stopped short of enabling users to select specific tracks or artists. The labels claimed that this feature set was “interactive” and therefore that LAUNCHcast should have to negotiate license terms rather than get a compulsory license with statutory royalties. The court found for LAUNCH Media. That decision paved the way for later customizable Internet radio services like Pandora and iHeartRadio, which were able to offer any recorded music track through compulsory licenses and pay statutory royalty rates.

There was also a loophole in sound recording copyrights for digital radio that remained for many years afterwards. The DPRA established copyright protection for digital broadcasts of sound recordings, but only as far back as 1972—because the Sound Recording Act of 1971 did not make it retroactive to older recordings.  Sound recordings made before February 1972 were subject to a patchwork of state copyright laws. This meant that digital radio services technically could not rely on statutory licenses and needed to negotiate licenses with labels for playing much of the music featured on oldies, classic rock, jazz, and classical stations—though of course this hardly ever happened. A series of lawsuits were filed by the likes of Flo & Eddie (lead singers of the Turtles as well as solo artists) and ABS Entertainment (label of Sam Cooke, Jackie Wilson, and other classic artists).

This loophole was finally closed in 2018 with the passage of the Music Modernization Act, although only for digital radio. The MMA was primarily focused on fixes to mechanical licensing of musical compositions for interactive streaming. But it also included a provision that was originally called the CLASSICS Act (Compensating Legacy Artists for their Songs, Service, and Important Contributions to Society), which established a performance right in pre-1972 sound recordings for digital radio. The final loophole wasn’t closed: there is still no performance right in sound recordings (of any vintage) for terrestrial radio.  The United States continues to be one of only a handful of countries in the world without this right; others are Iran, North Korea, and Rwanda.[3]

The more recent arguments over whether sound recordings should carry performance rights have turned on whether radio still has the promotional value for recorded music that it did in the pre-digital era. Whenever this issue comes up inside the Beltway, both sides—the RIAA for record labels, the NAB for broadcasters—unleash battalions of economists armed with studies purporting to show that it does or doesn’t. But the real reason why there is still no general performance right in sound recordings has to do with lobbying at the state level. And that takes us from copyright law to the FCC.

Radio frequencies are a limited resource, like land and water; so the government has regulated their use through the FCC (originally the Federal Radio Commission), which was established with the Communications Act of 1934. As radio stations proliferated around the country, the FCC moved to establish ownership limits: a single company could not own or operate more than seven stations nationwide, and no more than one in any given market. Although ownership caps were eventually relaxed in the 1980s (along with requirements for public service programming and other regulations), they helped shape a market for radio stations that had no nationally dominant players. National ownership caps were eliminated with the Communications Act of 1996. But by that time no more new frequencies were available, so if you wanted to own a radio station, you had to buy the license from someone else, which cost into the tens of millions for powerful stations in major markets.

When the 1996 Act passed, the big broadcast chains went on buying sprees. By the mid-2000s, four chains—Clear Channel, Citadel, Viacom, and Cumulus—had an aggregate dominant share of audience and revenue, and the total number of station owners had decreased by one-third. But even then, the total number of stations that those four chains owned was less than one-fifth of the total number of FCC-licensed stations. Even at its peak in 2003, Clear Channel, the largest chain, owned around 1200 stations out of about 15,000. (Today iHeartMedia—Clear Channel under a new name—owns about 850.)

As a result, most broadcast stations are owned by small to medium sized businesses that are based all over the country. In contrast, the recorded music industry is concentrated heavily in three states: New York (NYC), Tennessee (Nashville), and California (Los Angeles, Santa Monica, Burbank). This means that whenever Congress considers legislation that pits the interests of radio against those of the music industry, there are members of Congress in three states being lobbied by the RIAA, while members of Congress in most of the other states are being lobbied by the NAB. NAB’s nationwide reach gives them the clout to portray any royalty as a “performance tax,” and raising taxes tends to be anathema for many politicians. The NAB also garners political mileage by characterizing the royalties as a harm to small businesses and “small town radio.” That’s why there is no performance right in sound recordings today, and radio pays no royalties to labels or artists.

The clash between the labels and the broadcasting industry over this issue continues to this day. The last serious attempt to pass performance rights in sound recordings was the Fair Play Fair Pay Act of 2018, which the labels attempted to bundle into the MMA but had to jettison due to pressure from the NAB. Congress continues to consider various pieces of copyright reform legislation; the U.S. Copyright Office in early 2021 weighed in with a statement supporting a performance right in sound recordings, as it has done in the past.

[1] A new fifth PRO called Pro Music Rights, which licenses a catalog of mostly hip-hop compositions, has been recognized by some radio broadcasters.

[2] This set of regulations is known as the “DMCA performance complement,” referring to the Digital Millennium Copyright Act of 1998 in which the regulations were recodified (meaning that they received new section numbering in the copyright statute). The regulations were quietly abandoned in 2009 by agreement between the NAB and record labels.

[3] China was on this list until quite recently. It just passed legislation providing for royalties for radio play of recorded music, which went into effect on June 1, 2021.

[i] Hearings on Economic Conditions in the Performing Arts Before the Select Subcommittee on Education of the House Committee on Education and Labor, 87th Cong., 1st and 2d Sess. (1961-62).

[iii] Performance Rights in Sound Recordings, Subcommittee on Courts, Civil Liberties, and the Administration of Justice of the Committee of the Judiciary, House of Representatives, Ninety-Fifth Congress, June 1978.

Now here we are in 2021, where a technology exists that, one could argue, has some similarity of purpose but less functionality than those DRM schemes: it guarantees the authenticity of an object for the purchaser while not ensuring its uniqueness or scarcity. This technology is the latest hype magnet in tech; it is the enabling technology for a piece of digital art that sold at auction last week for $69 million. I’m talking, of course, about non-fungible tokens (NFTs): records stored on blockchains that link purchases to identities of purchasers.

What do NFTs do that DRM doesn’t do? Not much. To me, the only interesting thing is that NFT schemes make public the identities of people who paid money for the NFTs that point to digital objects, so that those people have help from the NFT platform in exercising bragging rights. DRM technologies generally don’t do this, though they certainly could.

Notice that in the previous paragraph I used the unwieldy phrase “paid money for the NFTs that point to digital objects” instead of the more straightforward “bought the digital objects.” That’s because NFTs, like many DRM schemes, create digital simulacra of ownership of creative works; and just like DRMs, NFTs fall short of emulating true ownership. Ownership implies scarcity; both NFTs and DRMs are attempts to impose scarcity on a technological world that resists it strenuously.

The winner of that auction at Christie’s last week, who goes by the pseudonym Metakovan, no more owns that pile of bits by the digital artist Beeple than I own the ebooks that I got from Amazon. I can read the ebooks on my laptop and my phone (and my Kindle reader, if I could find it in my closet), but I couldn’t read them on a Nook or Kobo (if I had either). I can’t resell, give away, or rent them; even if I were to hack the DRM, I’d be violating Amazon’s terms of service.

Metakovan can view the Beeple artwork on any device that will display JPEGs. But in a sense, he can’t alienate the work either, because it’s already available to everyone in exactly the same format as it is to him. He can’t lend or rent the NFT, because the NFT platform doesn’t support those operations (though it could). He can resell the NFT, but then he doesn’t get to keep all the revenue from the resale–commissions go back to both the artist and the NFT platform and/or auction house. (Resale royalties to artists for physical art objects are mandated by law in many countries, but not in the U.S.) It’s not clear whether he can give the NFT away (i.e., resell it for $0) or leave it to his children to inherit.

In neither the NFT nor DRM cases do buyers get the same bundle of rights that are guaranteed for a physical object in copyright law. In both cases a single private entity controls and benefits from the process in perpetuity, whereas no private entity is involved–or is even allowed by law to be involved–after someone purchases a physical copyrighted work.

There have also been integrations of NFTs with DRM. We’ve discussed a couple of them here: ebook schemes that use DRM along with NFTs for identification of ownership. This is essentially the same as a traditional DRM scheme, except that the transaction records and identifiers sit on a blockchain instead of in some vendor’s private database. These schemes support alienation (resale, giveaway, loan), but standard DRM schemes could do this too; for example, Amazon’s Kindle supports e-loans in limited circumstances, and both Adobe’s DRM and Readium LCP support library e-lending . Because blockchains are inherently ownerless, some of these providers have touted the prospect that blockchain-enabled ebook markets can be vendor-independent; so far nobody has figured out how to make the DRM part vendor-independent (although this may be possible with Readium LCP DRM).

It’s useful to speculate on the future of NFTs for content by looking at the trajectory of DRM since it first appeared. DRM still exists for ebooks in most places and for commercial TV shows and movies. It was tried and rejected for music downloads, but it’s still used for streaming music services. DRM was never used for digital art, mainly because digital art didn’t really exist during the era of DRM’s original technological development; it was never really used for images at all.

To the extent that NFTs for content make sense at all, they only do so for types of content that are unique objects–paintings, drawings, sculptures–or nearly unique objects–signed and numbered photographs or lithographs. This has to do with consumer expectations. Digital art, unlike digital music or video, imports consumer expectations from the physical world, and those expectations include uniqueness or near-uniqueness.

In contrast, the idea of an NFT for an article or song seems silly. The most prominent example of the latter, from the rock band Kings of Leon, is only plausible because it’s also tied to physical items like concert seats and limited-edition vinyl. On the other hand, an NFT for a piece of art is very much like a provenance record; it’s equally applicable to digital and physical artworks. Auction houses like Christie’s have always kept provenance records; blockchains are a convenient way to do this that didn’t exist until recently.

Yet consumer expectations change over time. Music is a good example of this. Streaming music services eschew music ownership entirely. Streaming was a niche curiosity in the early 2000s, but now–according to annual industry revenue numbers that the RIAA released last month–it accounts for over 80% of industry revenue, and the vast bulk of that comes from interactive services like Spotify that let users choose whatever they want to listen to. The only other segment of the music industry that’s growing is vinyl, which experienced an impressive 23% growth last year. In other words: after many years of being offered ersatz ownership of digital music via downloads, consumers have decided that they don’t care about ownership when it comes to digital music, and if they want to really own music, they’ll buy vinyl instead of MP3s (even DRM-free ones).

There’s a risk that the same kind of thing will happen to digital art. There are many ways that the NFT bubble can deflate; here are just a few I can think of. Some artists will decide to try emulating signed and numbered photographs/lithographs by selling multiple NFTs for single works of art; this will inevitably backfire if they get too greedy and the number gets too high, i.e., more like traditional DRM where the number is unlimited in “sales” situations. (Notably, U.S. law recognizes 200 as the maximum number of copies a visual work can have and be “limited edition.”) NFT bragging rights will become less meaningful if there are millions of NFTs in general floating around. And then there’s the potential for undermining the original alleged purposes of NFTs, when third parties try to create “counterfeit” (or “pirate”?) NFTs for objects that already have them, are owned by someone else, or are in the public domain. (Policing this would be similar to policing unauthorized uploads on a content-sharing site like YouTube; perhaps a subject for another day.)

Ultimately, what NFTs and DRMs have in common, and in common with countless other technologies, is that both will find their niches after lots of hype, experimentation, and backlash. Yet there is one huge difference: DRMs only generated interest among copyright owners, not consumers, while NFTs have generated massive consumer hype. And hype is what’s driving the mania for NFTs right now. There’s no better evidence of this than the fact that Metakovan runs an NFT investment fund and benefits directly from all that hype. He invested his $69 million not so much in a visual artwork as in publicity for NFTs. And at least in the near term, his investment certainly paid off.

This brings up two interesting issues about copyright in podcasting. First is about paywalls and licensing of content to distributors. “The History of Sketch Comedy” is part of a trend of podcasts behind paywalls that has been growing quietly over the last couple of years and is, in my view, headed for an upward inflection point shortly. (I’ve been writing about this in Forbes.) Spotify has been acquiring podcasts such as the highly popular “The Joe Rogan Experience” and making them exclusive to Spotify subscribers; it has also been launching new shows, through publishers it owns such as Parcast and Gimlet, and making them exclusive too. Amazon launched a podcast service with exclusive shows bundled into its Spotify competitor, Amazon Music (and thus separate from Audible), late last year. And Apple is expected to launch its own separate paid subscription service for podcasts later this year.

The article in the Times mentions Audible but doesn’t say that “The History of Sketch Comedy” is behind a subscription paywall. The Times has been publishing podcast reviews at least since the start of the pandemic last year, but this is probably the first time that it’s reviewed a platform-exclusive podcast. Not saying that “The History of Sketch Comedy” is exclusive to Audible is analogous to not saying that a new TV series is on Hulu or Netflix or Peacock or Amazon Prime or Disney Plus or Discovery+ or etc., etc.

In the past, podcasting was a license-free zone. Producers put their podcasts out there, often paying hosting services to do it for them, and just let all of the many podcast apps list them in their directories. Ubiquity was the guiding principle. The idea that someone (such as an app developer or service provider) would need a license to copyrighted material was utterly alien.

But that’s changing. The first sign of change was in 2019, when a startup called Luminary raised $100 million in an attempt to launch a “Netflix for podcasts” paid subscription service. Luminary’s apps played free podcasts in addition to the couple dozen exclusives that it offered. But Luminary had various problems at launch, one of which was that it edited URLs out of podcast show notes, claiming “security concerns”–but many of those URLs were links to crowdfunding or sponsor pages. The backlash was swift: several major podcast publishers–including the Times, Spotify, Barstool Sports, and iHeartRadio–demanded that Luminary take their podcasts down from its service, DMCA-style, stating that Luminary did not have permission to offer them.

The next big sign of change was Spotify’s deal with Joe Rogan in May of last year. The comedy star had already built up a formidable audience on his own; Spotify reportedly paid him over $100 million in return for an exclusive license. That is, Spotify does not own “The Joe Rogan Experience” (as it does, say, Gimlet’s tech podcast “Reply All”) but instead licenses it. More deals of this type will surely come. Most of the major independent podcast publishing houses are being acquired by the likes of Spotify and Amazon. Therefore individual celebrity shows like Rogan’s are more likely to stand out as must-have content for podcast paywall builders, and those podcasts are more likely to be made exclusive via licensing deals than by outright acquisitions.

The second copyright-related issue that Keegan-Michael Key’s podcast brings up has to do with the use of third-party copyrighted material in podcasts. We’ve recently discussed the challenges of licensing music for use in podcasts and the different approaches that services like Spotify, Pandora, and iHeart are taking to it. Yet because the prodigiously talented Key is doing his own performances of classic radio and television sketch comedy material on “The History of Sketch Comedy” instead of using audio from the actual clips, he doesn’t have to clear the rights to them.

Or does he? A couple of startup companies are now saying yes.

Two startups, Spoken Giants and Word Collections, are saying that the text works underlying podcasts and other recorded spoken-word audio, such as comedy, are copyrighted works themselves and must be licensed where necessary. Key’s podcast includes new performances of the text of recorded comedy works, as opposed to the original recordings themselves; this implies that even if he doesn’t need to license sound recordings (or audio from TV or film), he may need to license the underlying text content. This situation could also come up in converse situations, when podcasters want to license their material to multiple entities, or in some instances where people want to use portions of podcast scripts for various purposes, if the material is in copyright and there’s no obvious fair use defense.

Digital Entertainment World will be hosting a virtual panel on this subject on Thursday, February 11, featuring the two startups’ CEOs, which should provide more insight into this new area. (Full disclosure: I’m an advisor to Spoken Giants.)

As podcasting becomes bigger and bigger business, these and other copyright issues will become more and more important. Some will lament (or may already have lamented) this exodus from the halcyon days of podcasting, when everything was free, fancy production wasn’t expected, and all the ads were read by podcast hosts; but such is the price of fame.

Publishers have been talking about the benefits of rights and royalties management for decades, and plenty of vendors have offered solutions. Some of these are standalone solutions from vendors like Fadel and RightsZone; others are modules of enterprise software suites for publishers from vendors such as Virtusales, Firebrand Technologies, knk Software, and Ingenta. There’s been plenty of anecdotal evidence of the benefits of these systems but precious little evidence of the financial return on investment from them.

To underscore this, BISG published the results of a survey three years ago that showed that the majority of publishers believed that they were missing out on revenue opportunities from rights licensing and that improved
rights management would lead to increased business opportunities with licensees. But the survey also found that only 40% of respondents planned to make meaningful investments in rights management in the ensuing 3-5 years.

This gap between general awareness of benefits and specificity required to motivate investment hasn’t changed much over the past two decades. I was involved in a roughly similar project with the Association of American Publishers’ Enabling Technologies Committee back in the mid-1990s–a project that led to the development of the Digital Object Identifier (DOI) standard but no meaningful increase in publishers’ rights management processes.

This new study is an attempt to remedy the lack of credible and useful estimates of rights management ROI—estimates that are quantitative, in actionable detail, and from a sufficiently unbiased source. Our interviewees spanned trade, STM, professional, and academic publishing as well as literary agencies; and they ranged in size from small to very large. Almost all of them had never before considered the quantitative financial benefits of their rights and royalties systems and processes in detail. We hope that this study will encourage publishing executives to take a fresh look at rights management in their organizations.

Here are a few of the study’s highlights:

• A midsize trade publisher was able to increase licensing revenue by $2.1 million over a period of 5 years.
• A large trade publisher’s system saves it around $3 million per year in costs related to permissions granting.
• A large professional publisher increased its licensing revenue by about $600,000 through analysis of data about prior licensing deals, enabling it to improve pricing and increase license renewals.
• A large STM and professional publisher has been able to increase translation rights revenue by 25%.
• A large educational publisher saves over $9 million per year through greater rights and permissions efficiency and reduction of licensing-in costs.
• A small literary agency has increased revenues by 10–15% by increasing foreign rights deals per year with the same staff.
• A university press saves the cost of 6 staff by using a cloud-based permissions licensing service.

In addition, we found that themes emerged from our interviews that go beyond those that we’ve heard over the past several years. The new themes include:

• Data analysis: the volume of permissions and subrights deals has increased to the point that it’s worthwhile using data analysis tools to optimize sales efforts and pricing.
• Capacity expansion instead of cost reduction: more and more publishers are starting to see benefits of rights process automation as enabling more licensing deals instead of reducing time and effort of processing them. And the financial benefits of the former are generally larger than the latter.
• Online licensing agencies: the volume of licensing deals, especially in non-trade publishing, has reached the point that it’s no longer feasible to process them in-house. The aggregate revenue that online licensing agencies like Copyright Clearance Center and PubMatch process is now high enough that these agencies have moved from “nice to haves” to critical services.

Spotify announced earlier this week that it is enabling users of its Anchor podcast publishing platform to use music–entire tracks from Spotify’s massive catalog–in their podcasts. This “Music, Meet Podcasts” campaign indicates not only the current sorry state of music licensing for podcast use but also Spotify’s apparent ability to circumvent the difficulties.

Podcasts have become extremely popular: the latest Edison Research/Triton Digital Infinite Dial survey, from just before the start of the pandemic, shows that over 100 million Americans age 12 and up listen to podcasts regularly. Podcast revenue is starting to reach significant levels, too: the Interactive Advertising Bureau estimates that 2020 U.S. ad revenues in podcasting will approach $1 billion. Total podcasting revenue is more than that because additional revenue comes in through sponsorships, crowdsourcing, and subscriptions. That’s not much compared to $15 billion for music, but podcast ad revenue is growing at almost triple the rate of music.

There are many interesting issues of content licensing around podcasts. On the one hand, with a small (yet growing) number of exceptions, podcast publishers make their podcasts available to podcast apps and users with no licensing at all–they are free to everyone without restrictions.

On the other hand, there isn’t much music in podcasts. Most of it is either specially commissioned for individual podcasts or gotten from libraries of pre-cleared production music such as PodcastMusic.com or from Creative Commons music libraries. That’s because there is no simple licensing mechanism for commercial music in podcasts.

Good background on this topic can be had from–what else?–a podcast. On his latest Musonomics podcast, Larry Miller–head of the Music Business program at NYU, where I teach–discusses this in detail with several guests. Here’s a brief explanation.

Licensing recorded music requires licensing both sound recordings and the musical compositions they embody. Licensing sound recordings requires negotiation with record labels, who can refuse to license. Music publishers are generally required by law to license musical compositions for reproduction and distribution (mechanical licenses). But many music publishers consider the use of music in podcasts to require sync licenses–the same type of license used for music with video, such as movies, TV shows, or TV commercials. (Others in the music publishing industry disagree with that position because podcasts don’t contain video.) There is no simple mechanism for obtaining sync licenses, and music publishers can refuse to offer them at all.

So, podcasters have to negotiate with both record labels and music publishers to license each piece of recorded music for podcasts (unless they want to rely on fair use arguments). Leaving aside the fees that licensors charge, this is a grossly inefficient process. There is no blanket license available for podcasts, as there is in (for example) radio broadcasting from PROs like ASCAP and BMI. Nor is there a law that obligates copyright owners to make the material available (compulsory license), as there is in mechanical licenses for music downloading or streaming through agencies like the Harry Fox Agency.

Two of the guests on Larry Miller’s podcast express different views on this. Music clearance expert Deborah Mannis-Gardner says that music publishers aren’t motivated to create a simple mechanism for licensing because they don’t see enough revenue opportunities in the near term. On the other hand, music supervisor Scott Velazquez says that music publishers are currently discussing how to make the process easier. He predicts that it will be possible to license most commercial music easily and cheaply for podcasts, with the exception of “premium” music (songs by big-name artists or songwriters).

A possible solution to this problem would be to create a special category of sync license that delineates podcasting from more typical video use cases, such as “sync for audio only.” Unlike performance and reproduction rights, sync rights aren’t defined in copyright law; they are derivations of statutory rights that are defined by industry convention. Therefore there’s no reason why the industry can’t set up “sync for audio only” licenses also as a matter of convention.

The National Association of Music Publishers and/or the RIAA could set up an agency to manage these licenses, as the NMPA did with Harry Fox for mechanical licenses back in the 1920s, or as the RIAA did with SoundExchange to handle sound recording licensing for digital radio in 2000. (Both Harry Fox and SoundExchange subsequently became independent organizations.) Podcast publishers might have to form an organization to negotiate collectively for rates, as the Radio Music Licensing Committee does for performance royalties on broadcast radio or as DiMA does for mechanicals on streaming music services.

It will take a while to sort this out. In the meantime, the lack of easy music licensing mechanisms in podcasting leads, for example, to iHeartRadio not offering its syndicated American Top 40 radio show in podcast form, even though it offers dozens of other syndicated broadcast radio shows as podcasts.

Spotify is one of the few major podcasting platforms that also has full music licensing. This raises an interesting question: do Spotify’s existing music licenses (for reproduction, distribution, and public performance) cover podcast cases where rights holders would claim that sync licenses are needed? Spotify’s position is that they do, as long as the podcasts are hosted on its Anchor platform. It’s not clear whether Spotify arrived at this position unilaterally or after discussion with rights holders.

In any case, only paying Spotify subscribers can hear songs used in podcasts; they aren’t available to Spotify’s free subscribers. It’s also not clear whether Spotify put that restriction in place for licensing reasons–Spotify has different license terms with record labels for free vs. paid subscribers–or because Spotify wants another way to steer free users to paid subscriptions.

Other major music services have greater or lesser degrees of podcast publishing capability. Apple and Google have podcast apps, but neither of them offer podcast publishing or hosting services. Amazon recently launched a podcast feature within Amazon Music that features a few exclusive podcasts as well as all of the freely available ones; but it doesn’t offer podcast publishing or hosting tools per se.

The other major music services that are also significant podcast publishers are Pandora and iHeartRadio. Last year, Pandora launched a feature called Stories, which enables artists to create spoken-word pieces around playlists of their music, and enables third-party podcasters to use music from a selection of pre-cleared tracks.

In other words, Pandora is taking a more conservative view than Spotify on whether its existing music licenses cover the podcasting case. One reason for this is that Pandora Stories are accessible to all Pandora users, the vast majority of whom use its free service, which is only licensed for “non-interactive” (radio-like) and not on-demand music playing; and podcasting is an inherently on-demand medium.

As for iHeartRadio, it publishes several popular podcasts about music, but they don’t include much music in them. iHeartRadio is similar to Pandora in that the vast majority of its users use a free service that is only licensed for non-interactive music use.

Apart from these exceptions, the lack of commercial recorded music in podcasts is a big hole. This is especially true for those (like presumably iHeartRadio, as well as NPR) who believe that podcasting is the future of radio. The longer podcasts go without straightforward access to commercial music, the more established practice it will be to rely on free or low-cost production music, and the more lost opportunity there will be for music rights holders. Spotify’s new campaign–assuming it doesn’t raise legal objections from music publishers–gives it an advantage over its competitors in terms of the content it can offer in podcasts for its paying subscribers. It’s yet another example of Spotify’s strategic focus on podcasts. But that advantage shouldn’t last forever.

This announcement has gotten some notice in copyright circles; in the United States, the most common reaction has been, “Why?” Yet while WIPO PROOF isn’t valuable for U.S. copyright purposes, it’s valuable elsewhere in the world–and it’s valuable outside of the copyright realm.

WIPO PROOF works by using a few standard cryptography functions. One of these is hashes. A hash (a/k/a hash value) is the output of a special type of mathematical function that reduces arbitrarily large amounts of information to small fixed-size data, like an extreme type of data compression algorithm. It does this in such a way that two different inputs to the function are almost certain to produce different hash values even if they are only very slightly different, while two identical inputs will always produce the same hash value. The result is a number that serves as an identifier for content or data that’s unique for most practical purposes.

For example, if you input a file containing a mystery novel ebook to a hash function, you get a hash value X. If you input an exact copy of the ebook file to the same hash function, you will get X again. But then if you change one word of the novel–e.g., change “The butler did it” to “The cook did it”–you get a completely different hash value. (This isn’t the same as a fingerprint, which is a special type of hash value that denotes uniqueness of content as perceived by a human, so that, for example, two music files containing the same sound recording in different formats will produce the same fingerprint.)

Given a digital file as input, WIPO PROOF gets a hash value for the file from your browser (using the standard SHA-2 algorithm) and sends it to a server at WIPO. WIPO creates a timestamp, appends it to the hash, and creates a digital signature of the result using its highly secure private key.

A digital signature of a message–in this case the content file hash and timestamp–is a way to enable the message to be checked for tampering. Digital signatures also use hashes. A digital signature algorithm computes the hash of a message and encrypts the hash using a private key. If someone wants to ensure that the message hasn’t been altered, they can use the public key that corresponds to the private key to decrypt the signature, then use the same hash function to recompute the hash of the message. The decrypted signature and the recomputed hash should be equal; if they aren’t, then the message has been tampered with.

WIPO provides the public key needed to check the content hash and timestamp for integrity. It does this by adding a public key certificate. A public key certificate is essentially a way to have a trusted third party called a certificate authority vouch for the integrity and ownership of some data, in this case WIPO’s public key. WIPO uses certSIGN of Romania as its certificate authority. The public key certificate essentially says that certSIGN vouches for the fact that the public key belongs to WIPO and that WIPO is a trustworthy organization.

The content hash, timestamp, digital signature, and public key certificate all come together to form the WIPO PROOF token, a file that is just a few kilobytes in size. Once WIPO generates the token, the user can download it from WIPO’s website. WIPO keeps a copy in its database, as a backup and so that it can re-generate the token in the future if any of the cryptographic algorithms used to generate it are cracked.

All of these cryptographic functions have been in wide use for a long time. Yet this scheme has certain advantages. The most important is that it uses web browsers’ built-in standard hash functions to compute hashes. This means that the data that you want to timestamp never has to leave the computer it’s on, which is important for efficiency (the data could be very large) as well as confidentiality. Although web browsers have supported hash functions for a long time, the standard they supported until recently was SHA-1, which was hacked back in 2017; now browsers support the successor standard SHA-2.

Once you have a WIPO PROOF token, you can use it as evidence that you had possession of a file at a point in time. You do this by using the public key to decrypt the digital signature, as described above, to check the integrity of the content hash and timestamp. (WIPO provides an online tool that does this.) Then you can recompute the hash of the content and check to see that it matches the content hash in the token, which is proof that your file is the same one as (or an exact copy of) the file that you had at the time of the timestamp.

So what is the application of WIPO PROOF to copyright?

International copyright law provides that a creative work is copyrighted automatically as soon as it’s fixed in a tangible medium (such as a digital file) without the creator having to take any action. Proof of existence of a creative work at a certain point in time matters in litigation. Yet in U.S. copyright law, it’s necessary to register your work with the U.S. Copyright Office in order to sue for infringement, and your registration is your proof of existence at a point in time. The Supreme Court affirmed this just last year with its decision in Fourth Estate v. Wall-Street.com, which established that it isn’t sufficient merely to apply for copyright registration in order to sue for copyright infringement, the Copyright Office has to approve the application (a process that can take months). 

In this light, WIPO PROOF is merely a high-tech version of “poor man’s copyright,” such as sending your content to yourself in the mail and having the U.S. Postal Service’s postmark serve as a trusted timestamp. This has no value in U.S. law. At least one recent startup company has tried building blockchain versions of poor man’s copyright, where deposit on a blockchain serves as an unalterable record of ownership and creation date. These are equally worthless without actual Copyright registration.

However, the U.S. is the only country with this type of system. Several other countries have voluntary registration systems in which registration isn’t required to bring suit but can serve as evidence that a work existed at a point in time. WIPO PROOF is most useful in the many other countries that have no copyright registration formalities at all, or have antiquated, inefficient, paper-based systems. It is a simple and inexpensive way of bridging the digital divide in such places.

WIPO also sees WIPO PROOF as valuable in other types of IP management and protection scenarios. For example, data sets (such as in pharmaceutics and life sciences) typically don’t qualify for copyright or patent protection, yet it can be important to prove their ownership and origin dates. More generally, WIPO sees WIPO PROOF as a way to help people and businesses protect intellectual property at earlier points in their development before the need for formal protection sets in. That’s part of WIPO’s mission and the real value of WIPO PROOF.