Tag: Copyright

Korea’s AI Action Plan: Declaring War on Creators?

A young woman in a sparkling silver outfit poses next to a large robotic figure, adorned with South Korean symbols and colors, in a vibrant city background.

Image: Shutterstock

In the scramble to jump on the global AI bandwagon, Korea has floated a proposal that would supposedly remove “legal uncertainty” for AI developers who use copyrighted content to train AI platforms. Unfortunately, the proposed “solution” threatens to throw Korea’s globally renowned creative sector under the bus. Nor does it remove the uncertainty.

As part of President Lee Jae-Myung’s National Artificial Intelligence Strategy, its Presidential Council has put forward a 98 point “Action Plan”, a blueprint for implementation. There are many aspects to an AI strategy, but a key element is to ensure legal clarity with respect to the use of content for AI training, especially copyrighted content. The Action Plan purports to do this. Its Point 32 proposes the introduction of “explicit exceptions under the Copyright Act to allow copyrighted works to be used without legal uncertainty (emphasis added) in the processes of collecting and analyzing data available on the web”. In other words, introduction of a copyright exception for text and data mining (TDM), subject to certain conditions such as some form of remuneration, transparency, and opt-out features for rightsholders.

If the goal is to remove “legal uncertainty” regarding the use of copyrighted works, this proposal falls short of the mark. No exception can provide 100% certainty given the Berne Convention requirement that any exception meet the so-called “three step test”, meaning that an exception is permitted only in certain special cases, provided that it does not conflict with normal exploitation of the work and does not unreasonably prejudice the author’s legitimate interests. While there will always be a degree of uncertainty regarding exceptions, the good news is there is a ready alternative. The surest way to ensure legal certainty is to encourage licensing of content from rightsholders. The problem with the introduction of a TDM exception—or even the discussion of a possible TDM loophole–is that it diminishes the likelihood of reaching licensing solutions by reducing the pressure on AI developers to open their wallets to reach licensing deals.

It is even more bizarre that the tech industry is pressing for a TDM exception given that Korea is one of the few countries, alongside the United States, that has adopted a fair use provision in its Copyright law. This was done in 2011 as part of the implementation of the Korea-US Free Trade Agreement after heavy lobbying by the tech sector. Fair use allows courts to make case-by-case judgements as to whether a given use meets fair use criteria, thus potentially allowing reproduction of copyrighted material without advance permission from the rightsholder. If free use of copyrighted material for AI training can be shown to be “fair”, why is a TDM exception needed? Even in countries where fair use does not apply (which is most of the world), there is no convincing case or consensus on the need for a TDM exception; there is even less reason for one in a state that has already adopted fair use.

Point 32 of the Action Plan takes note of the existence of fair use in Korea, commenting that the Ministry of Culture, Sports and Tourism is preparing fair use guidelines. These are to provide interpretive guidance on the exemption provisions of the Copyright Act to enable companies to utilize copyrighted data “with greater confidence”. Despite this, the Action Plan claims these guidelines alone are unlikely to fully eliminate uncertainty and judicial risks. Voluntary licensing, however, would eliminate both.

It is well established that AI developers need vast amounts of data to improve the performance of their AI platforms. To date they have largely employed a “take first, ask later” policy. This has led to numerous lawsuits pitting rightsholders against the tech industry, mostly but not exclusively in the US. AI developers in the US have argued that what they are doing amounts to “fair use” because the final AI product is used for a different purpose from the original and thus does not compete with it. That is highly debatable, especially with image and music-based AI works. To date, the results from the US courts have been mixed.

The legality of the tech industry’s unauthorized use of copyrighted content is an issue that a number of countries, in Asia and around the world, are looking at. Various solutions have been proposed to eliminate the uncertainties that arise from leaving the decision to the courts. Among these are TDM exceptions which have been introduced, albeit with strict limitations, in the UK, the EU and Japan. In the UK for example, use is limited to non-commercial purposes. In the EU, it must be accompanied by transparency requirements and opt-out provisions for authors. In Japan, if the unauthorized user derives commercial benefit from the content, the safe harbour does not apply. Australia has explicitly ruled out introducing a TDM exception in order to protect its creative sector, while many others (eg. India, Canada) have no TDM provision in their copyright law. As noted above, the clearest way to remove any uncertainty about the legality of using copyrighted works is to incentivize and recognize voluntary licensing as the solution. This ensures that rightsholders receive appropriate compensation for the work they have put into creating content, while guaranteeing legal certainty for licensees.

The Korean strategy paper argues that AI companies are required to obtain individual consent from each copyright holder “leading to significant costs and time burdens in securing high-quality training data”. But large amounts of high-quality content can be accessed through voluntary licensing agreements with major content creation companies such as studios, publishers, broadcasters, music labels and so on. As for individual authors and artists, one possibility is to look at the model currently used for licensing print and music content through Collective Rights Management Organizations as a supplement to voluntary licenses signed with major rightsholders.

In addition to being instructed to prepare the necessary amendment to the Copyright Law for presentation to the National Assembly by Q2 of this year, the Culture, Sports and Tourism Ministry, in cooperation with the Ministry of Science and Technology, is to “promulgate standard contract templates for the licensing and transfer of copyrights for AI training”. This is the kind of heavy-handed market intervention that is guaranteed to stifle voluntary licensing. Not only that, it amounts to expropriating the rights of Korean creators to manage their works.

The Action Plan gives a nod to the importance of compensating rightsholders and claims it wants to establish a system that respects the rights of creators. However, given the size and importance of Korea’s cultural industries, from film to K-Pop to literature, it is surprising there isn’t greater recognition of what an important strategic and economic asset this sector represents for Korea. Although the Strategy acknowledges that content industries should be able to share the benefits of growth in the AI industry, the proposed solution is unbalanced and biased toward clearing any so-called “obstacles” to unimpeded use of content. As a result,  just days after the extremely brief (20 day) consultation period on the Strategy had closed, in mid-January sixteen creator and rightsholder groups issued a strong statement condemning the Action Plan, labelling it “an attempt to fundamentally undermine copyright as a private property right”.

While paying lip service to creator’s rights, the Plan does not address how creators can enforce these rights (other than through the creation of opt-out protocols, which stands the normal copyright procedure of seeking permission prior to usage on its head). The Strategy seems to lead to what has been described by many as a “use now, pay later” system, with little information on how payment would be calculated or implemented. On the other hand, prior, voluntary licensing of content for AI training is a solution that would respect the rights of Korea’s creators while providing the welcome revenue sharing and income stream for which the Strategy advocates. Strong content industries benefit AI development in Korea by encouraging continued creation of the valuable Korean language content so necessary to refine and improve AI models. Conversely, providing the tech industry with an escape hatch to avoid licensing by instituting a TDM exception is the surest way to kill a licensing market for AI content. It will only continue the legal uncertainty that the Presidential Council seems to feel is hindering AI development in Korea.

The one-sided formulation of the Strategy to date has provoked an inevitable negative reaction from Korea’s cultural industries. This is not surprising since the strategy of the tech industry, in Korea and elsewhere, is to avoid dealing with ministries directly responsible for culture and copyright and instead lobby industry, technology and science ministries to bring pressure for changes to copyright law. This adversarial stance is unfortunate as the content and tech industries need and can help each other. The Strategy needs to be amended so rather than throwing the cultural and copyright industries under the bus in the name of facilitating AI development, Korea provides the framework for a mutually beneficial and legally certain relationship. This is best done by upholding longstanding copyright principles and encouraging the growth of a voluntary licensing market for content used in AI training.

© Hugh Stephens, 2026. All Rights Reserved

Chinese (or is it Lunar?) New Year and Copyright

A festive design featuring a red background with gold and floral decorations, celebrating the Happy New Year 2026, themed around the Year of the Horse.

Image: Shutterstock

The China Media Group, (CMG) one of the world’s largest broadcasters, has just issued an announcement warning everyone, but mostly I suppose the Chinese-speaking world, that it retains the sole authority to license globally the mascots, logo, and creative products of its 2026 Spring Festival Gala, for both commercial and non-commercial purposes. According to this website, the Gala program which is broadcast each year in China on the eve of the Lunar New Year, “has been recognized by Guinness World Records as the most-watched annual television program on the planet”. Even bigger than the Super Bowl! This year, the Year of “Fire Horse”, the Gala will feature “four majestic ponies”. These equine mascots are apparently, “inspired by classic representations of horses across different periods of Chinese history” and are “adorned with classic elements of traditional Chinese clouds and thunder”. You’ve been warned. Don’t copy them. If you want to watch all the action, you can view the show here. (Check out the dancing warrior robots. If that doesn’t send a shiver down your spine, nothing will).

While using horse images to depict the Year of the Horse cannot, of course, be copyrighted, the CMG will have copyright on the specific horse designs it is using for its mascots. Lunar New Year (or should it be called “Chinese New Year”?) is big business, as hundreds of millions of people will be celebrating it. Gold coins and bars will be struck for collectors and countries around the world, from the US to the Isle of Man, will be issuing Chinese New Year animal-themed postage stamps. (The Hong Kong Postal Museum has a hilarious post outlining the political machinations surrounding the Year of the Pig postage stamp issued in 1970 by the British colonial authorities). From broadcasts to banquets, from music to mascots, the Sinosphere (and by that I also include Chinatowns around North America and the world) will go into a red-hued frenzy this year beginning on New Year’s Eve, February 16, and continuing for 16 days thereafter. The holiday period in China marks one of the most intense short term urban-to-rural migrations of people anywhere, with family members traditionally returning home to visit parents, children and other relatives and to perform traditional ceremonies. Rooted in tradition, this migration sees about 300 million people move from one end of China to the other. Rail networks in particular are strained to the limit. Visit China at some other time if you want to go there.

Entrepreneurial merchants will be quick to market equine-based trinkets and paraphernalia this year, and knockoffs of the CMG “majestic ponies” will no doubt be popular in flea markets around China. So just as Hasbro protects the copyright and trademarks on its “My Little Pony” franchise, CMG is sending warning that it will assert its rights, as it should.

Music is another area where copyright could come into play as people create New Year’s videos. This website (Soundstripe) has made a virtue of a necessity by plugging its royalty-free songs and playlists for “Festive Lunar New Year Videos”. And while we are on the topic, should the holiday be called “Lunar New Year” or “Chinese New Year”? It is of course based on the lunar calendar but also tied to the centuries old Chinese zodiac, with its 12 year animal based cycle.

The problem with the “Chinese New Year” (CNY) label, widely used in North America, is that it imposes a Chinese label on similar celebrations in other parts of East Asia, notably Korea (where it’s called Seollal) and Vietnam (where it’s called Tết), that also celebrate the holiday. China historically had a huge cultural influence over both countries which explains the common tradition, but in this day of political correctness (not to mention the Sino-Vietnam war in 1979 and ongoing difficult political relations between South Korea and China), it can be impolitic to refer to the holiday as “Chinese”. Korean-Australian K-pop artist Danielle Marsh found out the hard way when she asked her fans what they were doing for “Chinese” New Years. She had to apologize for the faux pas. This is all somewhat of a tempest in a teapot because its not even called Chinese New Year in China, where the holiday is referred to as “Spring Festival” (chun jie).

As noted, this is the Year of the (Fire) Horse, which is supposed to indicate resilience, progress and, of course prosperity. The zodiac is a goldmine for fortune-tellers and geomancers who are happy (for a fee) to provide advice to individuals based on their personal sign as to how to comport themselves in the Year of the Horse, Dragon, Snake, Monkey or whatever. Some animals are more propitious than others. Dragons are considered to be an elite sign, and who wants to be a Pig? However, all have their good and bad characteristics. For prospective Chinese couples, figuring out the compatibility of say, a Rat and a Rabbit, can be important not to mention how the Year of the Horse will affect that relationship.

Having been born in the Year of the Monkey, I was eager to find out what was in store for me in 2026, since horses and monkeys don’t seem to have much in common. I was relieved to learn from this zodiac website that “Monkeys enjoy romance and wealth opportunities in 2026, though career stress tests patience. With resilience and clear financial management, challenges turn into steady progress.” That is reassuring although I am not sure that my wife will be thrilled to hear of my “romance opportunities”. Wealth maybe.

Not much of this has a lot to do with copyright, but it is fun to write about traditions and festivals, and CNY is particularly fertile when it comes to images, music, food, design and other creative endeavours that make up the world of copyright. So, enjoy it. Hang up some red lanterns, put up red decals of harmonious characters, and eat some dumplings. However, if you are inclined to create some Year of the Horse images, make sure you don’t infringe the designs and mascots of the China Media Group. They have a long arm and just might come after you.

© Hugh Stephens, 2026. All Rights Reserved.  

Libraries Are Complaining About the Cost of Licensing E-Books (Again): Is There Any Justification?

An e-book reader displaying the text 'E-BOOK What book you will read today?' next to a stack of hardcover books and an open book.

Image: Shutterstock.com

Is it a coincidence that a story complaining about the cost of e-books in the budget of the Greater Victoria Public Library (GVPL) appeared in my local paper, the Times-Colonist, earlier this month only to be followed mere days later by an almost identical story that popped up in a news scan service I receive, but this time relating to the Washington, DC, Public Library? I suspect not. In the case of the GVPL, the library has to shell out $57.00 for a two-year licence for an e-book that costs $14.99 for an individual licence. Both the Canadian Library Association and its US counterpart, the American Library Association, have had this issue on their agenda for a few years now. The Canadian Urban Libraries Council, which represents more than fifty of the largest public library systems in Canada, has a Digital Content Working Group that advocates for “equitable, affordable, and sustainable access to digital content for public libraries across Canada.” Who could argue with that? But there is always a devil hiding in the detail.

The major difference between the two stories was that the GVPL was simply lamenting the budgetary impact of licensing e-books, which are in popular demand from borrowers, whereas the US story not only raised the cost issue for libraries but also referred to a local legislative initiative that would authorize market intervention by imposing restrictions on the District’s public library system, preventing it from licensing e-books from publishers that reportedly charge “excessive prices”. But what is excessive and what is reasonable when it comes to e-books, and can or should governments intervene in the market to regulate the price of a discretionary product like a book?

The DC legislation has not been adopted and if it passes, it may still never go into effect since it contains a proviso that implementation will take place only once ten other jurisdictions with a combined population of at least 50 million have enacted substantially similar laws. This is apparently to protect the DC Council from lawsuits and to try to build a national coalition to negotiate collectively with publishers. It also avoids the very real possibility that publishers could opt to not license e-books to the District of Columbia Public Library system. The bill would actually hurt libraries and their patrons by prohibiting the DC library system from entering into and renewing licensing agreements with publishers if those agreements contain terms that “restrict public access to books.” What does “restricting public access” mean? Are publishers not legally entitled to determine the terms under which their product is offered to various categories of users? Should publishers be legally required to license their product on terms arbitrarily set by government? Most licensing agreements, which are contracts, impose conditions governing how a licensed product can be used by different categories of users, such as (for a library) the number of times it can be loaned and how long the agreement lasts before it lapses and must be renewed. These terms are set by the entity offering the product, i.e. the publisher. There is no obligation on any prospective licensee to accept the terms. Products that are priced beyond what the market will accept will not succeed. The argument of the library community seems to be that libraries are somehow different, a public good, and therefore they should pay what individual consumers pay even though their use of the product is quite different.

The outline of the DC bill states that restricting public access includes “inflating” costs beyond what the public pays. Under standard licensing terms for digital books offered to the public, aka Joe and Jill Consumer, they pay less than libraries do for access to the same work. This has been the case for e-books since the early days of digital content, and it is not difficult to understand why. When a library buys a hard copy book, which it can do for the same price as the consumer (in fact, usually for less since library systems buy in volume and thus enjoy discounts), it can expect to lend that book for a limited period before it wears out and a replacement has to be purchased. There are other risks as well when hard copy works are circulated on loan, such as loss and damage. Moreover, while he majority of borrowers (in Victoria it is 55%) still generally prefer to borrow a hard copy version, there is “friction” in borrowing a physical book that limits the number of times a work is taken out. This friction is the unavoidable time and cost of physically accessing a library, such as parking or transit costs, deterrence from inclement weather or just sheer laziness. E-books, on the other hand, can be accessed with a click of a mouse. Each copy is perfect. No dog-eared or missing pages here. No borrowing “friction”.

While “friction” is an undesirable byproduct of the traditional library system, imposing constraints on borrowing physical products for some, from a commercial perspective it imposes a reasonable limit on use. The absence of “friction” where digital products are concerned not only makes things easier for the borrower, but it also reduces costs in other areas for libraries. E-books require virtually no physical storage space, and while some clients will go to a library to access an e-book, many do so remotely, thus reducing physical visitorship and the need for extended hours, etc. One of the key arguments put forward by publishers to justify the higher cost of e-book licences to libraries is that individual consumers cannot circulate digital copies to others (unless they physically lend the device on which their e-book is loaded) whereas lending to multiple users is the essential function of a library. An e-book licensed by a library for lending to multiple users is, in effect, a different product from the single-use product licensed to an individual consumer. Moreover the “perfect-every-time” perennial nature of an e-book eliminates the need for any book replacement over time. Combined with potential savings gained in other areas, the pleas of public libraries to have legislators impose restrictions on publishers have to be seen in context. That said, I am sure this will not stop library associations from trying to convince legislators to tilt negotiations in their favour.

The DC Council is modelling its legislation on a similar law passed last year in Connecticut. Like DC’s draft law, the Connecticut legislation will come into effect only when another state or states totalling seven million in population enact similar laws. For the record, Connecticut has a population of about 3.6 million people. Neighbouring New York state has a population of 20 million so if it goes the same route, the ball will start rolling. The caution of both DC and smaller jurisdictions like Connecticut is understandable because one response from publishers to legislation that restricts their ability to offer industry-standard contracts/licensing agreements could be non-availability of e-books for libraries in the affected areas. Libraries need to be careful what they wish for, since a heavy-handed approach by legislators could backfire.

Connecticut claims its legislation avoids the legal pitfalls that befell a 2022 law passed in Maryland that was blocked, on appeal from the American Association of Publishers and the Authors Guild, because the state law impinged on US federal copyright law. That state law required authors and publishers holding the rights to an e-book title to, among other things, offer unlimited copies of that title to public libraries in the state at an undetermined “reasonable price” when the title was offered to individual consumers. In effect it was tantamount to imposing a compulsory license on rightsholders and was rightly blocked by the courts. I wrote a long blog post on the e-book licensing situation at that time. You can read it here.

While the Maryland approach was overreach, the issue hasn’t gone away. Library budgets are under stress but that should not be used as an excuse to force subsidization from authors and publishers. As mentioned above, e-books never wear out and a widely circulated “frictionless” library e-book is not a direct substitute for the hard copy work that libraries used to buy–and still buy but in lesser amounts. It is a different product and to compare it to a hard copy book is to ignore commercial realities. While libraries are a public good, they still operate in the marketplace and affect the market for books, in some ways positively and in others negatively. Governments normally put their thumb on the scale of commercial negotiations only when there is market failure. This is far from the case today. The case for market intervention by government is weak, although I sympathize with the plight of acquisition librarians facing stretched budgets and trying to meet the increasing demand for e-books.

There are market limits to what publishers can charge libraries for e-books just there are market limits regarding what public libraries can pay. What is the appropriate licensing fee for an e-book configured for library use versus one sold to the individual reader? Is a multiple of three or four justified, or should it be less (or more?). An objective study would be helpful. Let’s hope there is room for negotiation that will fairly compensate authors and publishers while allowing public libraries to meet the growing demand of their clients for more digital works.

© Hugh Stephens 2026. All Rights Reserved

Canadians (and Anyone Else Outside the US): Beware the Annual Public Domain Hype

A black and white cartoon character resembling a mouse, wearing a hat and shorts, happily steering a ship's wheel.

Image: Public Domain

This is the time of year (the days and weeks after January 1) when, on a quiet news day, lazy journalists in Canada used to pick up and amplify a US based story about such and such a work falling out of copyright and into the public domain and write a story about it, complete with a grabbing headline, often to the effect that Mickey Mouse or Batman or The Great Gatsby or whoever is now “liberated from the chains of copyright”. The CBC did exactly that in 2024, producing a radio special on Steamboat Willie entering the public domain, completely ignoring the fact that all works created by Walt Disney and Ub Iwerks (which include the earliest editions of Steamboat) had already entered the public domain in Canada two years earlier, on January 1, 2022. Iwerks was co-author and joint rightsholder along with Disney for this work and as the co-author who lived the longest (he died in 1971; Disney in 1966) the term of copyright protection in Canada was based on the year he died plus, at the time, an additional 50 years. Thus, January 1, 2022 in Canada. As I wrote a couple of years ago (Canada is not the United States when it comes to Copyright: The Cases of Anne of Green Gables and Steamboat Willie (or Down the Copyright Rabbithole, Twice), sometimes works still under copyright protection in Canada are in the public domain in the US, and sometimes it is the reverse. Don’t assume. This law firm’s blog post (Gowlings) provides a good overview of what to watch out for.

The mistake of making the assumption that what happens in the US is automatically applicable to Canada is the unfortunate reality of being a cultural minnow living cheek by jowl with a content creation whale. This year I didn’t notice any of the reflected US public domain stories in the Canadian media, perhaps because the penny has finally dropped that public domain day in Canada will be a non-event until the year 2043 (no need for any hype), owing to the extension of Canada’s term of copyright protection from the life of the author plus 50 years to life plus 70 (for most works). Any works that had fallen into the public domain under Canada’s previous “life plus 50” term did not receive the additional term of protection but any works still copyright protected in Canada at the end of 2022 got another twenty years coverage before entering the public domain. Had Ub Iwerks died in 1972 instead of a year earlier, Steamboat Willie would have enjoyed another two decades of protection in Canada beyond what applies in the US, although it is doubtful whether the Walt Disney Company would have tried to enforce its rights under Canadian law.

Not only has Canada harmonized its current term of copyright protection with the US, EU, UK, and a number of other countries, (although there will always be discrepancies between the terms of protection afforded works in Canada versus those in the US for many years to come owing to the historical peculiarities of how the term of protection is calculated under US copyright law), there have also been fewer quiet news days in early 2026 thanks to the daily Donald Trump Reality Show. Moreover, there has been a surge in Canadian nationalism (and thus a greater awareness of cultural differences) as a result of the Donald’s 51st state taunts. (Prime Minister Mark Carney, an internationally recognized banker and financial executive seemed initially to enjoy Trump’s respect but since his Davos speech calling out the realities of the new world order, Carney, like his predecessor Justin Trudeau, has been demoted to the title of “Governor Carney” on Truth Social, apparently the current official channel for announcements of US government policy). So, journalists, if you want to write about what makes Canada different from the US, in addition to measuring distance in kilometers and saying “sorry” every time someone bumps into you, you could note that US and Canadian copyright laws are different. Similar in intent but not identical. For example, the US fair use doctrine with its unpredictable focus on transformative use does not apply in Canada, the US requirement for formal registration of copyright in order to bring legal action does not apply in Canada and, in particular, the complex (because of its convoluted history) US determination of when a particular work falls into the public domain does not apply in Canada.

This year, as it does every year, the Center for the Study of the Public Domain at Duke University’s Law School, published its Public Domain Day blog, highlighting all the works that fell into the US public domain on January 1, 2026. These include such well known works or characters as Agatha Christie’s The Murder at the Vicarage (protected in Canada until January 1, 2043) , Somerset Maugham’s Cakes and Ale (in the public domain in Canada since 2016), Blondie and Dagwood, nine additional Mickey Mouse cartoons, Dutch artist Piet Mondrian’s Composition No. II/Composition in Red, Blue, and Yellow, four songs by Ira and George Gershwin, and so on. Much is made of the fact that these works will be free to anyone to use, remix, copy and exploit but it’s not as if these works have been locked away in a closet, although their unlicensed use has been protected by copyright law. Copyright protection does not stop anyone from creating new works and while there may be limitations on hijacking Inspector Poirot there is nothing stopping aspiring writers from creating detective novels. There is another element worth noting as well. Particularly when it comes to copyrighted characters and cartoons, their later iterations may still be copyright protected in the US (because of the US baseline being date of publication plus a set number of years) not to mention protection offered by registered trademarks. Sometimes estates try to hang on to copyright protection at all costs, as appears to be happening in the US with Mondrian’s work. (It has been in the public domain in Canada since 1995, 50 years after Mondrian’s passing).

Once a work has entered the public domain it can be used in derivative works without permission. Has this resulted in a slate of new and creative works being produced for the benefit of mankind? Hardly. The usual result is for a brief surge of “edgy” productions incorporating a new public domain work, such as Steamboat Willie doing or saying things that Dear Old Uncle Walt (Disney) would never have countenanced. As I noted a couple of years ago, the “liberation” of copyright protected works has led to such triumphs as The Gay Gatsby, The Great Gatsby Undead (Zombie Edition) and the film Winnie the Pooh: Blood and Honey. So much for the public domain unleashing the juices of creativity.

For better or worse, copyright is not a perpetual property right. I support reasonable limitations on copyright protection including making a provision for works to enter the public domain after their prime exploitability has passed. This will vary by work with some works remaining evergreen, encouraging new investment into derivative works, updates, new editions, as well as providing ongoing returns to the estates of authors. However, at times there are situations where a work is long out of print and the rights-holder cannot be located, blocking a reprint. These situations can be dealt with through specific exceptions, much as fair use and fair dealing allow for specified unauthorized uses that do not damage the rights of the author.

To come back to the narrative that the public domain liberates content from the “shackles of copyright”, I contend this is nonsense. Beware the hype. And if you reside outside the US, don’t believe everything you read in the media regarding what works are in the public domain. You might be pleasantly surprised to find that a work has been in the public domain in your country for years (while Conan Doyle’s later works only entered the US public domain in 2023, they have been in the public domain in Canada since 1981). On the other hand, you might find the work you thought was free for adaptation based on what Duke University’s Center for the Study of the Public Domain says is way off base and it is still protected by copyright in your country of residence. Beware the hype and do your homework.

© Hugh Stephens, 2026. All Rights Reserved.

India’s Proposed Compulsory Licensing Scheme for AI Training: It Could be the Worst of All Worlds.

A wooden gavel striking a golden copyright symbol with the text 'COMPULSORY LICENSING' in bold.

Used with permission

In early December the Indian government through the Department for Promotion of Industry and Internal Trade (DPIIT), launched a public consultation on its proposal for a “One Nation, One Licence, One Payment” regime to govern AI training on copyrighted content. The comment period closes early next month. The stated intent is to establish a framework to allow AI developers to access proprietary content for AI algorithm training, without rightsholder authorization, while ostensibly taking into account the concerns and economic interests of those same rightsholders. The proposal from a semi-official committee established by DPIIT is based on a compulsory licence regime covering all content, past, present and future, with no opt outs for content owners. Given the unrepresentative nature of the committee, which failed to include rightsholders from many creative sectors, it is not surprising it has managed to come up with a solution that pleases absolutely no-one, not rightsholders nor the AI industry. Moreover, it is probably unworkable and could rightly be described as the “worst of all worlds”. How did we end up here?

It is widely accepted that to further refine and develop AI, more data is constantly required. Quality data is important. The better the data, the better the product. Many AI developers have, without authorization, already helped themselves to vast amounts of copyright protected material either through accessing pirate databases or by simply hoovering up publicly accessible (but nonetheless copyright protected) content on the internet. The result has been a series of lawsuits launched by rightsholders, primarily in the US but also in India, the UK and Canada. In the US the parameters of the fair use doctrine are being tested, with mixed results. There have been some settlements, notably the Bartz v Anthropic case in which AI developer Anthropic agreed to pay a group of authors $1.5 billion for having accessed and reproduced their works through pirate websites (and without authorization), as well as a large number of voluntary licensing agreements between corporate rightsholders such as media outlets, publishers and music labels, and AI companies. Clearly, legal leverage is needed to convince AI developers to negotiate with rightsholders (no one is volunteering, that’s for sure). In the US this is backed up by a statutory damages’ regime where the risk of being hit with large statutory damages per infraction is a strong incentive for the AI industry to reach licensing deals. Unfortunately, many countries do not have a statutory damages provision in their copyright law, and India is one of them.

Ironically, outside the US where the fair use case-by-case legal doctrine does not exist, AI developers are arguably in even greater legal jeopardy. As a result, they have pushed for a wide statutory text and data mining (TDM) exception to be introduced into national copyright laws. The AI industry is doing this in India, although there is strong opposition to introducing a TDM escape hatch. When a TDM exception applies, rightsholders are paid nothing. The DPIIT “solution” proposes to address the non-payment issue–but is going about it in precisely the wrong way. Appropriating the IP of all rightsholders by imposing a draconian compulsory license regime penalizes rather than rewards rightsholders and imposes a lowest-common-denominator value on all content, not distinguishing between premium and pedestrian content while taking away from rightsholders the inherent right to determine how and where their content is used.

The compulsory licence regime proposed by DPIIT would be administered by a new non-profit body, a Collective Management Organization (CMO) to be established by statute. This would create another layer of bureaucracy harking back to the days of the “licence raj”. A government appointed committee would set rates in conjunction with the new non-profit. Royalties would be applied retroactively as well as prospectively. Existing voluntary licensing agreements between content providers and AI companies would have to be terminated and future licence agreements prohibited. Rightsholders who do not want to licence their content could not opt-out. This is overreach at its worst.

The tech industry represented by NASSCOM, the National Association of Software and Service Companies, that also includes Google and Microsoft, vehemently opposes the One Nation, One Licence, One Payment proposal, arguing instead for a TDM exception. It also argues that rightsholders who wish to opt out should be able to do so, thus avoiding AI companies having to pay royalties for content it does not want or has already licensed. It also doesn’t want the precedent of compulsory payments, as India would be the first country globally to institute such a regime. NASSCOM claims that a compulsory licence regime would slow innovation, the card always played by the AI industry when the issue of protecting rightsholders is discussed. However, a TDM exception is not currently on offer in India. It is a “solution” that is facing increasing pushback in many countries. Australia has just rejected the concept, and other jurisdictions are examining it critically.  Where a TDM exception does exist, as in the UK and EU, its use is constrained and subject to several conditions, such as in Britain where it is limited to non-commercial research.

Not only is the tech industry strongly opposed to the compulsory licence proposal, so are rightsholders such as the broadcasting industry and Bollywood. Not only would a compulsory licence be extremely difficult to implement given the nature of CMOs which are generally highly inefficient and administratively cost heavy– not to mention that the Committee’s working paper proposes two layers of CMO (thus double handling and processing)– but the draconian and sweeping nature of the compulsory licence is of great concern. Rightsholders are given no ability to opt out or to refuse to have their content conscripted. In fact, the proposal includes a provision granting AI developers access to content as a “matter of right”. What happened to the fundamental right of an author to determine how, when and even if their content is to be reproduced? Compulsory licences are extraordinarily blunt instruments and do not work when sophisticated content like audio-visual and music products are involved. There are many elements to licensing, including contractual issues that go beyond price that are carefully negotiated and carry with them specific obligations and privileges. A compulsory licence is a “one size fits all” solution. It strips away the rights of content owners and is, in effect, a form of “compensated expropriation”. And the compensation is minimal.

Having pleased no-one with its proposal, it remains to be seen where DPIIT will go next once all comments are received and evaluated. A follow-up proposal is expected that will deal with outputs, just as the initial one did with inputs. India, like many countries, wants to participate in the global development of AI. Its rich local-language content is a strategic asset. Imposing a compulsory licence would stifle the creativity that drives this cultural comparative advantage. In the absence of a market failure there is no rationale for resorting to the sledgehammer of a compulsory licence for all content. The preferred solution for rightsholders, voluntary licensing, is also becoming a preferred solution for AI developers as the legal ground on which they are operating in accessing content without authorization is looking increasingly shaky. Voluntary licensing is growing globally as AI developers absorb this reality. If wide TDM exceptions are off the table, AI developers will have no recourse but to negotiate licence agreements. (A statutory damages regime would provide even greater incentive to do so). Bringing in a wide TDM exception or worse, introducing a blunt and highly bureaucratic compulsory licence regime, is a sure way to kill the growing voluntary licensing market.

DPIIT’s next steps will be crucial. If the goal is to create the conditions for a “made in India” AI industry while nurturing and protecting India’s valuable cultural assets there is no better solution than to create the conditions for the growth of a voluntary licensing market. Strong cultural industries and robust AI development go hand-in-hand. This means dispensing with forced, bureaucratic solutions like the proposed compulsory licensing regime while holding firm on rejecting a TDM loophole that would allow AI companies to plunder India’s cultural richness without any compensation to creators and rightsholders. Let’s hope India gets it right.

© Hugh Stephens, 2026.  All Rights Reserved

The Year That Was: Looking Back at Copyright Issues in 2025

A graphic featuring the text '2025 A Look Back' inside a magnifying glass on an orange background.

Image: Shutterstock.com

In last year’s annual retrospective, I commented that just about the only significant copyright story in town was the impact of AI on copyrighted content. There were three primary dimensions, unauthorized use of copyrighted content for AI training, copyright infringing outputs produced by AI platforms and whether works produced by or with AI qualify for copyright protection. The first two elements, especially the unauthorized use of copyrighted content for AI training, continues to be the big story in 2025. Little has changed in terms of the fundamental issues, although this year there were two key court rulings in the US that provided some legal guidance. These were the Bartz v Anthropic and Kadrey et al v META cases, both decided within hours in the same courthouse in San Francisco, but by different judges. As I noted in a blog post at the time, in July, the results were a very mixed bag. (Hold the Champagne: The Two AI Training/Copyright Decisions Released in the US Last Week Were a Mixed Bag for AI Developers). While both judges found the copying of copyrighted works to train AI algorithms “transformational”, and thus tending toward fair use, in the Anthropic case the AI developer was castigated for initially using two pirate data bases (Library Genesis, aka LibGen, and Pirate Library Mirror, aka PiLiMi), as source material. This use threatened to result in hundreds of millions of dollars in statutory damages for authors whose works were included in these databases (if they had registered their works with the US Copyright Office). Thus, it was with no great surprise that the world learned in September that Anthropic had agreed to a $1.5 billion settlement to bring an end to the case. (When the End Does Not Justify the Means: Anthropic’s $1.5 Billion Lesson). In the Kadrey case, the judge suggested the plaintiffs should have made a stronger market dilution argument, which might have overridden the transformation finding. (New AI works produced based on unauthorized inputs of copyrighted works will devalue and dilute the market for the original works). That theory has yet to be fully tested.

While the Anthropic settlement seemed large in absolute terms, many criticized it as just another cost of doing business that failed to protect authors. But that is what licensing is, a cost of doing business. The damages could have been larger than the settlement, so Anthropic made a strategic decision to settle. Given the precedent, META must be worried given the clear evidence that its top executives gave the green light to use LibGen for training purposes despite internal warnings about the risk of doing so. (Is it Ethical to Use Pirated Content for Commercial Purposes? META Thinks So). The Anthropic decision gave impetus to ongoing licensing discussions between AI developers and content owners, especially large corporate rightsholders, like news media, music labels and big studios, such as Disney. The list of companies striking significant licensing deals with AI developers, either for AI training or outputs containing derivative content, is rapidly growing. Meanwhile lawsuits, such as New York Times v OpenAI, continue in lieu of settlement agreements.

Outside the US where the “fair use” doctrine along with its transformation interpretation doesn’t apply, there are also lawsuits against AI developers (in Canada and India for example). However, in these “fair dealing” countries AI developers are in a potentially tighter spot. In many countries there is no statutory exception to allow unauthorized access to copyrighted content for AI training, commonly known as text and data mining (TDM). Where there is such an exception (as in the UK or EU), it is confined to very specific circumstances such as non-commercial research, or else it requires that rightsholders be provided with an opt-out mechanism. Given these inconvenient facts (and their inability to rely on fair use arguments outside the US), AI developers have been mounting a full-court press to have TDM exceptions introduced into national legislation using the pretexts that a) everyone is doing it (which is certainly not true) and b) if governments don’t create an exception allowing TDM for AI training, all those AI development funds will flow elsewhere. In other words, throw rightsholders and creative industries under the bus in order to chase some ephemeral AI research funds. Australia has just rejected this binary approach, announcing that a TDM exception will not be part of its review of copyright laws to help address the needs of the AI industry. A number of Asian countries are reviewing the need for a TDM exception, but are rightly being very cautious not wanting to sacrifice the vital cultural and economic interests their creative sectors represent. Canada is another country without a TDM exception, a situation that has made OpenAI jittery since they are being sued in Ontario Superior Court by a consortium of Canadian news publishers for copyright infringement, bypassing Technical Protection Measures (protected paywalls) and breach of contract. OpenAI has been trying to get the case moved to the US by challenging the jurisdiction of the Canadian court, so far unsuccessfully.

Just as lawsuits in the US are providing the impetus for settlement discussions between AI developers and rightsholders, lawsuits against the AI industry outside the US will potentially have the same effect. The big story of 2025 is how many licensing agreements have been already reached. Even META, which in Canada and more recently Australia insisted it doesn’t need news content and would not pay for it, has reached a news media licensing deal with a number of companies including USA Today, People, CNN, Fox News, The Daily Caller, Washington Examiner and Le Monde. Back in 2021 META agreed to license news content in Australia as a result of the introduction of the Australian News Media Bargaining Code, but when Canada enacted a similar provision, it refused to do so and evaded the obligation to pay for local news by blocking it on its Facebook and Instagram platforms. Subsequently, when the Australian Code came around for renewal this year, META balked. Australia is still contemplating its next steps and while it is treading carefully, indications are that it is prepared to move against META.

One reason why Australia is moving carefully is the Trump factor. Trump’s erratic behaviour including the tearing up of the established rules of international trade is the second big copyright theme of 2025, after AI’s predations and encroachments. Not that Donald Trump knows much about copyright or understands it, but in his broadsides against trading partners and the international trading system, copyright industries inevitably become caught up in his web. Whether it is threatening tariffs on movies filmed outside the US, regardless of the fact that the largest slice of the box office comes from non-US sources and many films are either co-financed by offshore producers or require non-US settings (Trump’s Threatened Tariffs on Hollywood Films Produced Outside the US: The Medicine Could be Worse than the Disease), or taking aim at the policies of other countries that may have a digital or content component, like the Digital Services Tax in Canada,and its Online News Act, Trump is unpredictable and often off-target. His initial tariff measures against Canada included everything from steel to autos, based on the specious “national security” argument that there was a flood of fentanyl coming into the US from north of the border (actually about 0.2% of the total in 2024), but so far he has not specifically targetted the cultural industries. That may all change as the negotiations for renewal of the CUSMA/USMCA begin in the new year and as Canada struggles to redefine “Canadian content”.

There are industry groups in the US that claim provisions of the Online Streaming Act violate the terms of the CUSMA/USMCA by discriminating against US content providers, and that the only way Canada can justify these provisions is to invoke the cultural exception clause of the CUSMA. This would allow the US to retaliate in any sector with equivalent commercial effect. I have questioned this interpretation but we will probably never know who is right because the Online Streaming measures are unlikely to be challenged by the US prior to the start of CUSMA/USMCA renegotiations in early 2026. In those negotiations, the US will almost certainly take aim at various elements of cultural legislation like Online News and Online Streaming, along with other issues like dairy supply management. These may or may not be subject to negotiation as part of the renewal of CUSMA, if indeed the US or Canada are prepared to renew it.

Even if Canadian actions are consistent with the terms of the Agreement, this is essentially meaningless given the way the Trump Administration operates, as demonstrated earlier this year when Trump overrode various tariff free commitments in the Agreement on the basis of unilaterally declared national security concerns. These included the specious fentanyl trafficking claim mentioned above. Not only did the miniscule numbers give the lie to this assertion but the US claim somehow made Canada responsible for ensuring the security of the US border. As professor Fen Hampson of Carleton University has eloquently pointed out, the fate of CUSMA/USMCA depends not so much on where the economic interests of the United States lie, but on the fickle and unpredictable whims of Donald Trump personally.

As we head into 2026, there will be more Donald Trump uncertainty affecting copyright industries, more AI disruption, more lawsuits, more settlements and more licensing agreements to avoid lawsuits and settlements. That’s a pretty safe prediction based on looking back on 2025.

© Hugh Stephens, 2025. All Rights Reserved.

The CRTC’s “Rube Goldberg” Definition of Canadian Content (CanCon): More Complicated..but Also More Flexible

An old television displaying the words 'CANCON REDEFINED' over a background of the Canadian flag.

Since it is frequently in the news, it’s worth asking the question. What is Canadian Content (Cancon)? It can be many things to many people. Unlike pornography, you don’t always know it when you see it. Blogger Michael Geist illustrated the problem well a few years ago with his Cancon quiz. If you want to do well in the quiz, select just about any production that the general public is likely to regard as Canadian– i.e. based on a book written by a Canadian, starring a prominent Canadian actor or notably taking place in Canada–as not qualifying as certified Cancon. Then select all the obscure productions you have never heard of including several with no identifiable connection to Canada as certified Canadian content. You will be a winner! This perverse outcome is because of the way the system is set up, as I have written about in previous blogs, such as this one (Unravelling the Complexities of the Canadian Content (Cancon) Conundrum).

In brief, up to now Cancon has been primarily defined by the number of points (out of 10) that a production accumulates, in addition to other factors such as the requirement that it be produced by a Canadian and reach a minimum 75% production expenditure in Canada (except for co-productions). Cancon is defined in regulation by no less than three entities, Telefilm Canada for co-productions, the Canadian Audio-Visual Certification Office (CAVCO), part of Heritage Canada, to determine eligibility for subsidies, and the broadcast regulator, the CRTC, with respect to meeting Cancon broadcast quotas. All use the points system, with some productions requiring 10/10 to obtain maximum subsidies, while most others  meeting a minimum 6/10 requirement. Points are awarded for the positions in the production filled by Canadians, such as the writer, director, performers, director of photography, production designer, music composer and picture editor. For CAVCO productions, the copyright must also be held by a Canadian producer for a minimum of 25 years. The actual story and its setting are completely irrelevant. In short, it is more of an industrial than a cultural policy, based on the assumption that if Canadians are in charge, they will produce content that reflects Canada. It often doesn’t work out that way.

Now the CRTC has updated its definition of Cancon as part of the implementation of the Online Streaming Act, which brings streaming services in Canada under the oversight of the broadcast regulator. Foreign streaming services over a certain revenue threshold are being required to make a financial contribution to Cancon (although they are challenging this in court) and may be required to promote Cancon on their services (“discoverabilty). The CRTC cannot impose broadcast quotas on an à la carte streaming service, whether domestic or foreign, thus the financial contribution and likely discoverability requirements. The survival and promotion of Canadian content, both domestically and internationally, is at the core of the legislation. Thus, the CRTC’s new definition of Cancon is very relevant.

If you thought the definition was going to get simpler, think again. However, it has been updated to incorporate new positions in productions, like a showrunner, plus those responsible for costume design, make-up artists, and hair artists. But not all productions, especially those in Québec, have all these positions, especially the new category of showrunner. As a compromise, having a Canadian showrunner will be worth an optional 2 bonus points, but if you don’t have one you won’t be penalized. What exactly is a showrunner? There is a lengthy CRTC definition related to the position being the creative leader of a production, managing the production process etc. With respect to costume design, make-up and hair artists, if collectively all these positions are filled by Canadians the production will garner another bonus point. If a production does not utilize all of these positions, it must fill the ones that it does with Canadians to get the optional point. Is Canadian makeup and hair design different from non-Canadian? I wouldn’t have thought so, but there you go. As I said, it is an industrial policy as much as a cultural one.

Here is another example of what is starting to look like very much like a Rube Goldberg machine, with add-ons, exceptions, secret doorways and special conditions. For animated productions, the Commission will now award 2 points (instead of 1 point) for each of the key creative positions Director, and Scriptwriter and Storyboard Supervisor, when filled by Canadians. There are various other tweaks; for animated productions, the Commission will award the points noted below for the following key creative positions, when filled by Canadians; Director (2 points; previously 1 point); Scriptwriter and Storyboard Supervisor (2 points; previously 1 point); and First Voice (or first lead performer) and Second Voice (or second lead performer) (1 point each; previously 1 point for one or the other, but not both). It goes on. For animated productions, the Director OR Scriptwriter and Storyboard Supervisor, and either the First Voice (or first lead performer) OR Second Voice (or second lead performer), and Key Animation AND Camera Operator must be Canadian. There’s more, adding Visual Effects Director and Special Effects Director to the list of key creative positions in a film, adding one bonus point if both are Canadian.

If all this has your head spinning, be assured that this stuff is of intrinsic interest to the industry but of not much relevance to Canadian consumers. What Canadian consumers want are Canadian stories in Canadian settings. On this score, there is a bit of a breakthrough, recognizing the importance of these factors for the first time. It is only a small opening but is the first time that location depicted in a film has been included as a factor in assessing Canadian content, as well as points for the source of the story.

The Commission will award 1 bonus point where identifiable Canadian characters and identifiable Canadian settings are included in a production, but all lead characters (up to 5 main fictional characters in dramatic productions) must be identified as Canadian or members of First Nations, Inuit or Métis in Canada and all persons on screen in non-dramatic productions (presenters, musicians, dancers) might likewise be Canadian, First Nations, Inuit or Métis (as if the latter were not, by definition, Canadian). As for location, “The location of the story must be set in Canada. The story or narrative must take place entirely in an identified Canadian city/region/province/territory. The location can be identified by a Canadian landmark or by identification on screen or otherwise identified overtly in the narrative or text of the program.”

All this for one lousy point! If you want to incorporate a visual reference to a place outside Canada (for example, one’s homeland for immigrant Canadians) could you do it in a dream scene if the dreamer’s bedroom has a shot of the CN Tower through the window. Not clear. But it is a start toward recognizing that settings, characters and stories are relevant to Cancon. A bonus point will be awarded for a production based on a Canadian story and another point for using Canadian music.

Finally, on the copyright front where the current CAVCO policy requires a Canadian to control the copyright for 25 years, there is mixed news. In a recent blog post, I argued that the CRTC should not impose a Canadian copyright restriction if the goal is to get foreign streamers to produce more Cancon for distribution abroad. Content is softpower. Content exported abroad not only helps cover the cost of production, it projects an image of Canada to the world through Canadian stories. To penalize foreign producers by preventing them from acquiring copyright in productions they have financed or partially financed, should they wish to acquire it, is shortsighted in my view. The CRTC took account of this concern but also had to listen to the instructions it received from government requiring it to consider the need to support Canadian ownership of intellectual property.

The end result is a compromise; Canadians must retain at least 20% of the copyright ownership in a program. In other words, up to 80% of the copyright in a production can be held by a foreign enterprise. In such cases, however, the production must accumulate at least 80% of possible points and the director and screenwriter must be Canadian. Where there are greater degrees of Canadian copyright ownership, some of these requirements are relaxed. There will be no minimum copyright retention period. A recent blog on MediaPolicy.ca goes into more detail on this.

Finally, there is the question of AI, just about the only point in the CRTC decision picked up by the mainstream media. The new positions created to increase the point count have to be staffed by humans, not AI. The rest of the CRTC package was likely too difficult to compress into something readable for the average news consumer.

What does this all add up to? An incredibly complex and bureaucratic system yet that is, believe it or not, a bit more flexible with respect to defining Canadian content than previously. It is a result of the classic compromises that must be made between idealism and reality, between promoting Canadian content in a bubble and ensuring its presence in the real, competitive world. It attempts to strike a balance between heavy lobbying by domestic constituencies such as Canadian independent producers and licensed broadcasters, and the foreign streamers that increasingly dominate the market. It tips the balance a bit more toward being a cultural than an industrial policy, but from the point of view of the average Canadian, is about as arcane as a bureaucratic process can get.

As noted, it is a Rube Goldberg machine with many levers needing to be pulled to get to the desired end, often by the most complicated route possible. But the Commission had little choice given that the current Cancon policy was clearly outdated. Maybe at the end of the day, we will actually get more recognizable Canadian content that finds audiences both domestically and internationally, on a variety of platforms. And while purist Canadian nationalists may disagree, if the new policy encourages additional investment in Cancon from the streamers, that can only be a benefit to Canada.

© Hugh Stephens, 2025. All Rights Reserved.

If anyone is not familiar with Rube Goldberg and his penchant for drawing overly complicated solutions to simple problems, this link will provide more detail.

We need more Canada in the Training Data, but through Licensing not Loopholes

Canada Has a Choice When it Comes to AI Training Content

Scrabble tiles arranged to display the words 'LOOPHOLES' and 'LICENSING' on a game board.

Michael Geist, Canada Research Chair in Internet and E-Commerce Law at the University of Ottawa has argued, in an appearance before the Heritage Committee of the House of Commons, that “we need more Canada in the training data”. He is absolutely right, but just not in the way he proposes. Dr. Geist is what I would call a well-known skeptic when it comes to the intrinsic value of copyright, a copyright “minimalist” if you will (probably an understatement).

With respect to the unauthorized and uncompensated use of copyrighted content for AI training, he states that “in the context of AI, the application of copyright isn’t clear cut. The outputs of AI systems rarely rise (to) the level of actual infringement given that the expression may be similar or inspired by another source, but it is not a direct copy of the original.” Whether the outputs mirror the inputs is not the sole issue. In some cases, such as when music and images have provided the inputs, they do. This is an infringement of the reproduction right, and likely also an infringement of the distribution right and the right to produce a derivative copy (under US law). In Canada the right to create another work from an original work comes from the right of adaptation. However, even without a mirrored output, full reproduction still takes place at the input stage, creating an infringement unless the copies meet a fair dealing purpose and fulfill fair dealing criteria, even if the copies are later deleted. As Keith Kupferschmid, CEO of the Washington DC based Copyright Alliance has pointed out in a recent blog post discussing the copyright principles that apply in AI training cases,

“Some people mistakenly believe that in order to establish an infringement during the input stage, the copyright owner needs to establish substantial similarity between the ingested copyrighted work and AI-generated output and if no substantial similarity exists there is no infringement in this stage. That is incorrect.”

Even without mirrored outputs, full non-transitory copies of copyrighted works are being made at the ingestion stage of AI training. That is an infringement, just as making a photocopy of a complete work, such as a book, would be an infringement unless covered by an explicit exception such as preservation purposes by a library or archive. 

Dr. Geist’s second line of argument is that if Canada makes it more difficult or costly to develop large language models, AI development will shift outside the country. This is a tried-and-true but tired pretext frequently employed by those seeking to justify the appropriation of copyright protected content in the name of “innovation”, as I pointed out in an earlier blog post. (CanLII v CasewayAI: Defendant Trots Out AI Industry’s Misinformation and Scare Tactics -But Don’t Panic, Canada). This is a race to the bottom, throwing the content industry under the bus on the pretext that everyone is doing it, even though that is untrue. One provision that has been selectively incorporated into the laws of some jurisdictions, like the UK and the EU, is an exception for “text and data mining” (TDM). Dr. Geist states this is why Canada also needs to introduce a similar statutory exception to promote AI.

However, not everyone is engaged in this race to the bottom. In fact, there are increasing doubts that establishing a statutory TDM exception for AI training is the best way to go. Australia has just firmly rejected the creation of a TDM exception in its copyright law even though it is also grappling with the same issue of how to incentivize AI training and research in that country. The UK’s current TDM exception is limited to non-commercial research purposes and in the face of strong opposition from its creative sector, Britain has put proposals to expand TDM on hold. Even the EU’s TDM law, which has two aspects, one limiting the data mining to non-commercial scientific research conducted by scientific research organizations or cultural heritage institutions while the other is a general purpose TDM that is open to commercial organizations, has guardrails. These include an opt-out provision whereby rightsholders can block ingestion of their content through technical measures, contract provisions or other means, in which case the TDM exception does not apply.

While opting-out by rightsholders is one way to limit the damage of unrestricted text and data mining, this is controversial because it places the onus on the rightsholder to take action whereas normally a party wanting to use someone else’s property would have to obtain permission in advance. Opting out is not a preferred solution for the creative community. It doesn’t work well in practice as rightsholders often lack the technical means or awareness to apply their opt-out rights. Because of this, the European Parliament’s Committee on Legal Affairs has just published a study examining how generative artificial intelligence interacts with European Union copyright law. The study recommends moving from opt-out to opt-in for rightsholders.

Thus, far from TDM being or becoming the norm, it is being rejected or constrained in a number of countries where the AI industry has been pushing it as the ultimate solution. The Canadian creative community, like the creative sector in Australia,  has spoken out strongly against introducing a TDM exception into Canadian law. Indeed, there is no need to do so as licensing solutions allowing AI training and text and data mining are becoming more and more common, including in Canada. For example, the Writers Union of Canada is studying a proposed agreement between select nonfiction authors, HarperCollins, and Microsoft to license full texts for the purpose of training artificial intelligence. Licensing agreements have taken off big-time in the US and elsewhere as the AI industry begins to understand this is the safest way to protect their investments. Canadian creators risk being left by the roadside if Canada brings in a TDM exception that would allow AI developers to steam ahead, appropriating content without payment or permission and ignoring licensing requirements by hiding behind a TDM exception.  The surest way to kill a nascent and growing licensing market is to give the AI sector a TDM loophole to exploit, removing any incentive to reach licensing agreements with rightsholders.  The solution is licensing, not loopholes.

Dr. Geist stated in his testimony to the Heritage Committee that AI developers would take the view that if they had to pay for (i.e. to license) content from Canadian creators, they would simply exclude it. The record of licensing deals being reached elsewhere suggests this is completely off base. Instead, the record shows that when AI developers want reliable, curated content to make their product better than the competition, they are ready to pay for it. But they will never pay for it if they are given a blank cheque through a legislated loophole. He also claims the position of the creative community is “Don’t use my stuff”. Again, the record of licensing deals to date and in the pipeline disproves this characterization in spades. Rather than blocking use of their content, creators are saying, “If you want to use my content, let’s talk”. Finally, Dr. Geist managed to completely mischaracterize the position of the creative community with regard to licensing. He said in his testimony that creators are advocating for a change to copyright law to mandate payments for AI training use. On the contrary, the creative community is simply asking that existing copyright law not be gutted. There is no need to create a mandatory payment requirement; existing copyright law is fit for purpose in dealing with how those wishing to use copyrighted content for purposes that fall outside fair dealing can do so. Negotiate a licence.

If any proof is needed of how the creation of a loophole will kill a licensing market is, all one needs to do is look at the sorry state of educational publishing in Canada. The industry has been decimated, and many authors have lost their livelihood because of the ill-conceived educational exception that was introduced into Canada’s Copyright Act in 2012. With that loophole in place, educational institutions across the country, with the notable exception of Quebec, began to tear up the reproduction licenses they had held from Access Copyright, the copyright collective representing authors. The educational exemption as part of fair dealing criteria could still be fixed, but the educational sector, facing severe financial pressures, has a powerful lobby working against it. The financial pressures are real, but taking a free ride on educational publishers and authors is wrong.

What happened with educational publishing is a cautionary tale for Canada. It should not make the same mistake twice. The way to promote a strong AI industry, alongside vibrant content industries, is licensing, not loopholes. Building a robust AI/TDM licensing market is the way to get more Canada into the training data, not giving the AI industry a blank cheque to help itself to the proprietorial content of others. With voluntary licensing everyone benefits. AI developers get secure access to quality content; the creative sector is rewarded for its efforts and becomes a partner in developing responsible AI. It’s a shame that the Canada Research Chair at the University of Ottawa doesn’t understand this.

© Hugh Stephens, 2025. All Rights Reserved.

The Online Streaming Act: Dealing with US Industry Concerns (The Cancon Factor)

A vintage television set displaying the Canadian flag.

Image: Shutterstock

Last month I wrote about the spectre of US trade retaliation against measures impacting or possibly impacting US streaming services as the Canadian Radio-Television and Telecommunications Commission (CRTC) proceeds with implementation of the Online Streaming Act (formerly Bill C-11). The Computer & Communications Industry Association (CCIA), a US trade association that includes, among others, Amazon, Google, Meta and Apple, has taken aim at this process, claiming it is discriminatory and violates Canadian commitments under the CUSMA/USMCA, the trade agreement that replaced NAFTA. A core element of CCIA’s argument is that the very concept of Canadian content (Cancon) is discriminatory because it violates Article 19.4 of CUSMA which calls for “national treatment” of a digital product;

“No Party shall accord less favorable treatment to a digital product created, produced, published, contracted for, commissioned, or first made available on commercial terms in the territory of another Party, or to a digital product of which the author, performer, producer, developer, or owner is a person of another Party, than it accords to other like digital products”

What is Article 19.4 intended to cover? As an example, if rules are imposed prohibiting digital products from causing harm to children, the same rules should apply to both domestic and foreign products. However, do special requirements regarding Cancon audiovisual or music products (such as airtime quotas on radio or possible “discoverability” requirements for streamers) constitute discrimination against US digital products? Maybe. Is all music and AV content fungible or is Cancon somehow different, i.e. not a “like digital product”? If Cancon is “different”, what is it that differentiates it? That is not an easy question to answer because of the many criteria that go into determining whether a product is considered Canadian for regulatory purposes.

I took a stab a couple of years ago at explaining how Cancon is defined (“Unravelling the Complexities of the Canadian Content (Cancon) Conundrum”). For AV products, it is basically a combination of four elements; production control, copyright and distribution rights, creative positions and production spend. The CRTC definition and the definition used by the Canadian Audio-Visual Certification Office (CAVCO), which distributes certain tax credits, are slightly different with the latter being more stringent. For music there is the MAPL system. As explained by the CRTC, to qualify as Canadian content, a musical selection must generally fulfil at least two of the following conditions: M (music): the music is composed entirely by a Canadian; A (artist): the music is, or the lyrics are, performed principally by a Canadian; P (performance): the musical selection consists of a live performance that is recorded wholly in Canada, or performed wholly in Canada and broadcast live in Canada, and L (lyrics): the lyrics are written entirely by a Canadian. The CRTC is proposing that the “P” criteria be dropped owing to changing patterns in the music industry, notably the many Canadian artists recording outside Canada, such as in Nashville.

Qualifying as Cancon is complicated, but it has value. Cancon certification provides access to various subsidies and funds as well as providing a product that meets airtime and broadcast obligations, where and when they exist. In the aftermath of the enactment of the Online Streaming legislation, a key question is whether streamers (like broadcasters) will be required to meet certain content quotas, if indeed it is even feasible to impose content quotas on streamers. The different delivery model, where it is the consumer who “pulls” content from a broad menu rather than a broadcaster who “serves up” a given offering, makes it almost impossible to impose content quotas. Theoretically, you could require a streamer to make available a specified inventory of Cancon, or even to promote Cancon (referred to as “discoverability”), but there is no way of making consumers actually watch or listen to Canadian productions. Trying to apply a 20th century broadcast model of regulation to 21st century streaming is not a good fit. Regulators around the world are grappling with this reality. One of the arguments for imposing an expenditure requirement on streamers, both domestic and international, to support the creation of Cancon is to compensate for the lack of applicability of content quotas in a streaming environment.

A core feature of certified Cancon at present is that it cannot, by definition, be produced by a non-Canadian regardless of whether all the creative talent (writers, directors, performers, designers, composers etc.) and production spend would otherwise meet Cancon criteria. There is a complicated formula that awards points for creative roles filled by Canadians, with a specified number of points required to qualify under different programs.  The fact that a non-Canadian production may be a Canadian story filmed in Canada with Canadian actors is irrelevant with respect to Cancon certification. In short, the colour of the money (the production company) is a determining factor. Additionally, under CAVCO rules, a foreign studio or producer cannot hold the intellectual property, (the copyright) in a Cancon production. A Canadian production company must be the copyright holder for all commercial exploitation purposes for a minimum of 25 years.

As part of implementing the Online Streaming Act, the CRTC was instructed to review the definition of Cancon. The Commission subsequently held public hearings in which ownership of copyright became a key issue. Opinions ranged from expanding the CAVCO requirements to all forms of Cancon to eliminating copyright ownership as a factor. The streamers, who now have (contested) financial obligations to fund Cancon, generally prefer to own copyright in productions. It is not a surprise that they object to being required to fund Cancon productions while being denied the opportunity to own and exploit the rights. Supporters of a more restrictive Cancon definition point out that foreign streamers are free to license Cancon qualifying productions from the Canadian rightsholder. However, a restrictive definition tied to financing and copyright ownership eliminates the possibility of direct financing by foreign streamers and could mean they would in effect be paying twice, first by contributing to the Fund that financed the production and second, by paying to acquire the rights. Moreover, there is no guarantee that the rights would be available on acceptable terms.

Those advocating for a comprehensive Cancon definition that includes financing and IP ownership as factors argue this is necessary to create and maintain a viable Canadian industry. But such restrictions have two effects. First, if copyrights must be retained, this removes from Canadian producers/rightsholders the ability to sell the rights at a time of their choosing (and possibly use the funds to produce more Cancon). Not all productions will have a sustaining revenue stream over time. It should be left to the producer to judge whether to cash out now or license the product while retaining ownership. Second, requiring that the producer be Canadian for a production to be certified as Cancon disincentivizes foreign streamers from self producing content showcasing Canadian stories, artists, locations etc. They can do so but are denied all Cancon credit for such productions. The cost of such productions does not count against their required financial contribution (currently 5% of revenues) nor does the production qualify as Canadian content in terms of meeting existing (or possibly future) content quotas. If a goal of Cancon policy is to promote expressions of Canadian culture through creation of financially viable productions, disincentivizing foreign producers from putting their toes into the Cancon lake makes no sense. Production of Cancon by global enterprises like the streamers will help ensure global distribution, meeting both cultural projection objectives as well as exposing Cancon to new markets.

There is also the question of subsidies provided to producers of Cancon. Under current definitions, the US studios are not eligible to access funds earmarked to produce Cancon (even though they are required to contribute to these funds). This could be dealt with giving foreign studios “contribution credit” for self-financed Cancon productions. It’s worth noting the studios are already offered generous subsidies–euphemistically referred to as tax credits–to undertake non-Canadian production in Canada, and no-one complains about that, except Donald Trump. Trump has been rattling the chains over so-called “runaway production” and has threatened to impose tariffs on movies made outside the US.

While I think many of the concerns of the foreign streamers could be addressed through a more flexible definition of Canadian content, I am not confident the CRTC will see it this way given the policy instructions it received from the government at the time the legislation was proclaimed. Can it comply with this guidance while not painting itself into a CUSMA corner? The Commission is directed to take international commitments into account, although there is no specific reference to CUSMA, only the 2005 UNESCO Convention on Cultural Diversity.

From my perspective it is not realistic for US streamers to expect a free ride (and they probably don’t) but Canada and the CRTC need to avoid being too greedy. They should also be flexible in defining Cancon, focussing more on the promotion of Canadian stories, music and talent and less on maintenance of an industrial policy that relies on protectionism for a favoured few. A policy that calls on foreign streamers to invest in Canadian creativity, given the revenues that they generate in Canada, is not unreasonable; denying them the ability to take a direct ownership stake in the products to which they contribute funding would be short-sighted. The policy straitjacket that exists with respect to Cancon sets up a search for draconian solutions, like the CCIA’s threats. In short, remove the Cancon handcuffs and keep the required contributions reasonable. Give credit for funds expended on content that meets Canadian artistic and cultural criteria. I think this would help blunt the frontal attack from US audiovisual streamers. Music is more complicated. Meanwhile, Canada needs to be careful not to negate any trade obligations it has taken on and avoid being forced into the Article 32.6 “cultural exemption” corner. 

But wait, I have an idea! If all else fails, there is also CUSMA Article 32.2 (b). “Nothing in this agreement shall be construed to…. preclude a Party from applying measures that it considers necessary for the fulfilment of its obligations with respect to …the protection of its own essential security interests.” If Donald Trump considers that importing kitchen cabinets from Canada threatens the national security of the United States perhaps it is not such a stretch to conclude that the preservation of Canadian culture (whatever that is) is just as essential to Canada’s national security, justifying any measures one chooses to employ. Is this a serious option? You decide.

© Hugh Stephens, 2025.  All Rights Reserved.

In writing this opinion piece, I have drawn on my background both as a former Canadian government official who has had some dealings with international trade issues over the years, as well as past experience as an executive with one of the US companies which, at the time, controlled a major Hollywood studio. (Time Warner). However, whatever “solutions” I have proposed to address US industry concerns regarding Cancon are mine alone. I hope they are a useful contribution to the debate, but I want to be clear that I do not speak for the CCIA or the streamers.

Australia Stands Up for its Creative Sector: A Useful Lesson for Canada and Others

Two coffee mugs side by side, one featuring the Australian flag and the other featuring the Canadian flag.

Image: Shutterstock

Australia just took an important stand in the tug-of-war being waged in many countries over whether, how and to what extent tech companies can use copyrighted content (text, music, images and so on) to train AI platforms by reproducing the content and extracting its essence without permission or compensation to rightsholders. Attorney-General Michelle Rowland has announced that while Australia will be undertaking consultations on revisions to its copyright laws to help address the needs of the AI industry, a Text and Data Mining (TDM) exception has been ruled out. Some countries, like the UK, have TDM exceptions for limited purposes (such as research and non-commercial use) in their laws while several other countries have TDM under review. Existing TDM exceptions allow reproduction of copyrighted content without the authorization of the rightsholder for research, data analysis, and in some cases for AI training purposes.

There is currently no TDM exception in Canadian law but as I noted in a recent blog post (“Canada’s Creative Sector Uneasily Awaits the Carney Government’s Next Steps on AI Training”), pressure is building from the AI sector to incorporate TDM into Canada’s Copyright Act. The government currently has yet another consultation paper on AI out for public comment and the Canadian cultural sector is organizing to protect creator’s rights, specifically calling on the Canadian government to “ensure that the Copyright Act is not modified through an exception permitting Text and Data Mining (TDM) or any other exception allowing technology developers or users to use protected works…to train generative AI systems without authorization or compensation…”. In doing so, it is taking a leaf from the book of Australian creators who mounted strong opposition to a proposal from the Productivity Commission, (PC) an independent research and advisory body created by an Act of Parliament some 25 years ago, that proposed in a report in August that Australia adopt a TDM exception. To say that this proposal put the cat amongst the pigeons would be an understatement.

The Commission has a reputation for denigrating the value of intellectual property and seeing it as an obstacle to industrial development rather than as an essential partner. In 2015 it proposed shortening the term of copyright protection from the current life of the author plus seventy years (“life plus 70”), a generally accepted international standard, to just “life plus 15”, (far lower than the Berne Convention minimum and a standard not adopted anywhere) while introducing a US-style fair use regime into Australia. There was strong pushback then, (it didn’t happen) and there was strong pushback this year (see here and here, for example) when the PC proposed introducing a TDM exception. It was particularly criticized for its lack of consultation with the creative industries in developing this proposal.

Now the Australian government has put its foot down, ruling out TDM but indicating that it will look at alternative solutions. These include examining whether to establish a new “paid collective licensing framework” under the Copyright Act for AI, or whether to maintain the status quo through voluntary licensing, clarifying how copyright law applies to material generated through the use of AI (i.e. whether there should be copyright protection for outputs produced by or with AI) and looking at the establishment of a new small claims forum to address lower-value copyright infringement matters.

It is generally accepted that AI is here to stay and will continue to need vast amounts of content for training. In most cases, copyrighted content is the kind of curated, high value work that AI developers need but until now, have preferred to appropriate without permission rather than pay for through licensing. In effect they have decided to ask for forgiveness after rather than permission beforehand. This has led to a plethora of lawsuits globally, including the recent $1.5 billion settlement that Anthropic has agreed to pay out to settle a class action suit brought by authors in the US. “Forgiveness” can be expensive. Inside the US, AI developers are arguing their copying is fair use, although at the same time they are beginning to hedge their bets by licensing content from a number of sources, ranging from media to music to image companies. Outside the US, AI companies have been beating the TDM drum, hoping that creation of wide TDM exceptions will obviate the need to negotiate with content owners. Nonetheless, voluntary licensing is growing globally. However, the surest way to kill a nascent licensing market is to give the tech industry a “get out of jail free” card by introducing a broad TDM exception. Australia has just rejected that option. Canada and others considering introducing new, or broadening existing, TDM loopholes should do the same.

It is not clear where Australia’s AI and Copyright review will end up, other than to note that it will not include TDM. As I have noted above, among other things it will be considering “collective licensing”. Collective licensing could help address the problem of remunerating individual rightsholders, in contrast to licence agreements signed between AI developers and corporate entities like media companies. However, Australia needs to steer clear of compulsory licensing which strips away the rights of copyright owners. Compulsory licences authorize use upon payment of a statutory or negotiated fee but remove the right of a copyright holder to withhold consent for use, or to impose specific limitations. A voluntary licence framework is fair to everyone. Compulsory licensing is not.

Canada and Australia have many things in common, (as well as a number of differences of course, beyond poutine vs vegemite). Among their commonalities is the desire to protect and foster a unique cultural identity in the face of global cultural homogenization. This is even more important in Canada given the realities of the struggle faced by 6 or 7 million Francophones to preserve their culture in a sea of 375 million Anglophones. Canada followed Australia’s lead (although less successfully) in requiring major online platforms to contribute financially to (i.e. pay for the use of) news media content. It should do the same by putting the idea of a TDM exception firmly to one side and instead focus on encouraging the development of voluntary licensing market for copyrighted content when used in AI training.

© Hugh Stephens, 2025. All Rights Reserved.