It seems as if almost every other day I receive another one, a somewhat convincing personal pitch to help me promote my 2023 book “In Defence of Copyright” to readers and thus potential purchasers. The pitches come from various sources but are almost always signed by a woman for some reason. The writer professes a strong interest in wanting to help me promote the book. “Amelia” (attractive picture), an “author success partner” was particularly committed to getting me on board. She talked about the many attributes of the book, its timeliness and potential and the importance of the topic. After she had immersed herself in my book, she began building a customized marketing strategy to help the book find its audience because, after all, “This book deserves to reach creators, policymakers, educators, and anyone concerned about the future of creative work in the digital age”. How flattering! What was this strategy and how much would it cost me?
A “Visibility Plan” was prepared that had a range of options, ranging from a few hundred dollars to a couple of thousand. At the basic level, the plan would “fix the storefront” by optimizing its profile on Amazon and Goodreads and inserting keywords and category positioning language. This “campaign” would last for 15 days. The next step, at roughly double the price, was to deal with “reach, momentum and carriage” would include Amazon KDP (Kindle Digital Publishing) advertising, Amazon best seller positioning and book club distribution. But the granddaddy was Option Three, at roughly double the price again, to include everything included in the first two levels plus Amazon short book trailers, author website design and social media management. Not surprisingly Option Three was recommended.
Another unsolicited pitch opened with the following, “Congratulations on the sharp insights, critical acclaim, and timely necessity of In Defence of Copyright, it is an extraordinary feat to craft a work of legal and cultural commentary with such a fierce, compelling hold on its readers.” Gosh, even more flattering. This service offered an eclectic range of services that included ”A bespoke, deep-dive narrative review sent directly to our community of dedicated readers”, a dedicated live online session focused on the book, an “exclusive, behind-the-scenes summary of anonymous reader insights on your book’s thematic impact”, what was called “C-Suite and Curator Introduction” which would apparently introduce the work to a private network of literary curators, book club syndicators and collection builders and, finally, “tailored promotional artwork” which could, if I wanted to pay the freight, include a 3D image of the book and long-term placement within a curated collection for the remainder of 2026.
What these pitches and a number of others I have received have in common is that they are specific to my book. They include multiple references to the content, themes, analysis, case studies etc. to the point that if one wanted to suspend judgement, you could almost be convinced that the people behind the promotion had actually read the book. But we all know that is not how it works. “In Defence of Copyright”, like hundreds of thousands of other copyrighted books, has been scraped and ingested by AI, without my knowledge or permission. The AI is then used to produce a reasonable summary of the book in support of the sales pitch. The website Writer Beware is a great source of information about the cruel world out there, particularly the blog post “Return of the Nigerian Prince Redux: Beware Book Club and Book Review Scams“ , which goes into some detail as to how the promotional offers actually work (or don’t). My experience with these promotions is fairly recent (perhaps I should be flattered that I have finally “joined the club”), but it has been going on for a couple of years, if not longer.
I am not saying that in some cases the services offered would not be provided. Possibly some promotional tweaking would actually take place, probably using some sort of automated system, but whether this would produce any noticeable bump in book sales, and whether the investment in these promotional efforts would pay off, is quite another question. The buyer has to decide this for themself. The business model employed by the book promoters is not dissimilar to automated cold calling with promotional offers; most recipients will simply ignore the messages, or they will go to spam. However, a certain number of people (me included, out of curiosity) will engage. An even smaller number will be converted to paying customers. However, if your initial outreach is AI driven and automated, even a one percent conversion rate can make the promotion worthwhile.
In my blog post last week, I talked about collective societies or CMOs (collective management organizations) as one possible solution to enable individual authors to have some control over how their works are used and to get compensated for that use. I believe that the “ART principles” (authorization, remuneration, and transparency) should apply, but this is not happening. Not only are authors getting no compensation for the use of their work in feeding the AI machine, but the unauthorized ingestion process is being weaponized against them by highlighting AI-driven analysis of their work to convince them to engage with automated promotion services.
The automation of the process can produce some amusing outcomes, however. Even though an AI platform may be adept at summarizing a work, it can still make rookie mistakes such as the pitch I recently received to promote “The Cambridge Companion to Gay and Lesbian Writing”, allegedly a work I had edited. Like the pitch on “In Defence of Copyright”, it contained flowery language about what an important work it was (in the field of queer literature), citing some of the key features of the work while pointing out how a bit of promotion would help it reach readers who really counted. A bit of digging revealed that this work was indeed edited by a namesake, Hugh Stevens, who is apparently an expert in this field and a senior lecturer at University College London. However, not only did the AI wrongly attribute Dr. Stevens’ work to me, but it couldn’t even differentiate between Hugh Stephens and Hugh Stevens, an annoying blind spot for many people that I have been fighting against all my life. Perhaps in this case the AI platform was not to blame; just the careless human being operating it?
Anyway, all I can say is that all the promotional spam I and other authors receive daily or weekly is annoying. Reaching out to book clubs and getting your work prominence on Goodreads are no doubt useful things to do, although in my case I don’t think many book clubs are interested in the ins and out of copyright (although a group of budding authors might find a quick overview of the topic to be useful). But I would submit that you can do a lot of your own marketing, with a good website and active outreach. While the spam is annoying, even more outrageous is the fact that it is a regular reminder to authors that their work has been copied and used to feed the AI machine without permission or compensation.
Is there a role for collective licensing and Collective Management Organizations (CMOs) in the licensing of AI, both for training (engaging the reproduction right) and use (distribution, publication, communication to the public and performance), and if so, what is that role? As the battle over AI’s use of copyrighted content for training and delivery of results continues globally, taking on different manifestations in different countries (some legal, some legislative, some consultative), rightsholders have argued that voluntary licensing is the best way forward. I argued just that point in a recent blog posting (AI Training and Copyright: It’s a Global Problem—and There is a Global Solution).
Voluntary licensing avoids the necessity for lawsuits, fair use or fair dealing determinations, and legislative amendments to change copyright law with all the unintended consequences that may entail. It also has the advantage of being multi-jurisdictional (assuming a broad licence is available), avoiding issues arising from different national iterations of copyright law, and provides both the tech and creative communities with a stake in AI development. It gives rightsholders control over the content they own or have created and, normally, a revenue stream. In some cases, such as licensing of prominent audio-visual repertoire, ongoing news content, or iconic images, that revenue flow could be substantial. In others, such as the licensing of an individual literary or musical work, the revenue might be minimal, but the issue of authorization would still be addressed. However, there is a fundamental issue associated with licensing the content of individual creators. In a practical sense, how do you do it? How do you reach all those individual or small rightsholders? This is where collective licensing could have a role to play. But it’s complicated (see here and here) and comes with problems as well as solutions.
Some AI developers, such as Anthropic in Australia, have proposed various options, such as an extended compulsory licence as the way to deal with the myriad of small rightsholders. The AI industry would pay into a central fund that would be used to compensate rightsholders who would be represented by a sectoral CMO, whether or not they were actually members of that organization. While a compulsory licensing scheme is one way to provide some revenue to small rightsholders, it removes one of the fundamental rights inherent in copyright, the author’s right to control the use of their work. Dealing with this problem requires rightsholders to opt-out, which is both technically difficult to implement and imposes a reverse-onus on rightsholders.
A fundamental principle of copyright law is that rightsholders have the right to authorize use of their work and should not have to exercise an opt-out proviso to prevent unauthorized use. Earlier this month a broad coalition of international news publishers wrote to Australian Prime Minister Anthony Albanese urging Australia to reject the kind of scheme that Anthropic is proposing, pointing out that such measures would be in violation of several international treaty commitments, would destroy the voluntary AI licensing markets that are rapidly developing and would establish Australia “as a location to launder infringement worldwide”.
In Canada, the umbrella cultural group Coalition for the Diversity of Cultural Expression (CDCE) has put forward three pillars governing the use of copyrighted content in the AI industry, which it calls the “ART principles” 1) authorization 2) remuneration and 3) transparency. Transparency is key because to date almost no-one outside the AI companies themselves knows exactly what the companies have ingested (reproduced) as part of their training, although it is far to assume “just about everything”. For many ingested works, the AI crawlers have absorbed the content on a “one and done” basis, that is they have used the content as cannon fodder as they scooped up vast quantities of data to train their bots. In other instances, involving use and re-use of the content, or what is called “Retrieval Augmented Generation”, the bot goes back to the original unauthorized reproduction of the work to extract information to provide answers to queries from users.
I tested this out with my own book “In Defence of Copyright“. I asked Google’s Gemini (I don’t subscribe to other AI services) a number of relatively obscure questions about the book. It had no difficulty whatsoever in providing quite detailed answers, and then asking me if I wanted more information, providing several suggestions as to what I could ask. To be able to respond in this manner, it was clear to me that the full work (not just the samples available through Amazon) must have been copied and ingested. No one asked my permission, I did not know the work had been copied, nor was I remunerated. None of the ART principles was followed. This story is repeated for probably just about every work under copyright ever published.
As I am a member of Access Copyright, the CMO for literary works in Canada, I wondered if it could help. Unlike its US and UK counterparts, Access Copyright does not offer a licence for AI training but is exploring the issuance of end-user licencing that would cover AI use. In the UK, the counterpart to Access, the Copyright Licensing Agency (CLA) announced the launch of an AI training licence last year, while the Copyright Clearance Center (CCC) in the US now offers an AI Systems Training Licence. Australia’s Copyright Agency does not have a licence for AI training but has extended its annual business licence to encompass AI uses. For example, this enables business licensees to allow staff to include third-party content in prompts in AI tools used in the workplace. The CLA and CCC also offer similar licences for use of AI by businesses.
Access Copyright told me it has chosen to focus on end-user licensing (rather than focusing on the AI companies themselves) because that is where it sees potential: business users feed copyright-protected content, owned by varied rightsholders, into AI platforms every day. These uses are dispersed and decentralized, unmonitored and unknown. A collective licence helps to alleviate the risks associated with these uses, by authorizing the use and remunerating the creators. In any case, since AI companies have already ingested most works available digitally (and are now embarked on copying and then pulping historical works available only in hard copy), it is questionable if there is really much of a market left for licensing literary works for AI training. That train may have already left the station. I am not planning on “In Defence of Copyright” generating any AI revenues for my retirement.
There is, however, another role for CMOs beyond offering licences for use of member’s works, and that is to enforce members’ rights. SOCAN, the Canadian music collective which, according to its website represents “over 200,000 songwriter, composer, and music publisher members” has just launched a lawsuit against music streaming service SUNO alleging that the company’s generative AI platform is “producing and streaming outputs that replicate human-created musical works without consent or payment, and as a result has infringed the performing rights in musical works in SOCAN’s repertoire”. In other words, the suit is focused on outputs, rather than inputs, as discussed here. It wants SUNO (and presumably other streaming services) to license content in its repertoire (plus pay damages of $10 million for the 150 infringed works). It also wants an injunction from the court requiring SUNO to take down the infringing works. The suit is similar to one that SOCAN’s German counterpart recently won against SUNO and is based on infringement of the performance right. Music is of course different from a published work, which may be read only once by a consumer while music tracks are consumed over and over again. There are also differences in the rights embedded in the works.
How and what impact collective licensing will have on the further development of AI, and the livelihood of creators, is still being explored. Rightsholders need to retain control. However, coming back to my initial question as to whether there is a role for collective licensing and for Collective Management Organizations (CMOs) in the licensing of AI, I submit that the answer is “Yes”. Figuring out exactly how to do it is the tough part.
This striking graphic, created by IP lawyer Maryna Hryhorieva, is a great summation of the current state of play regarding the still very fluid issue of AI training and copyright. It compares how eight different regimes, the US, EU, Japan, South Korea, China, the UK, Canada and Australia, are addressing this thorny issue, and highlights the different frameworks used to determine what is legal and what’s not. Each is different with its own peculiarities. Navigating through them is not easy. What may apply in one jurisdiction will not apply in another. A use that is acceptable in Country A may result in an unfavourable court ruling in Country B. An AI company can decide to base its content use practices on the most AI-friendly regime in which it operates, staying onside in that jurisdiction, but then can still be found to be infringing elsewhere. Cross-border jeopardy is almost certain where an AI company wants to scoop up local cultural content, pitting it against determined rightsholders who will, among other things, invoke cultural sovereignty.
The essence of the issue goes back to the way in which international copyright law is structured. Each country sets its own rules, regulations and interpretations within a broad commonly agreed framework. That framework, based on the Berne Convention, results in common basic standards centred on some fundamental principles, incarnated in the famous “three step test”.[i] The result is a generally consistent pattern of protection for rightsholders but with some significant differences by jurisdiction, like duration of copyright protection for example. The same inconsistencies apply to data mining (or what I would call unauthorized content reproduction) for AI training. So, what is a poor little ol’ trillion-dollar AI developer to do? Guess what? There is a universally applicable solution. Can you guess what it is? I reveal all in the final paragraph.
First, let’s briefly summarize the chart.
The US approach is based squarely on litigation, letting the courts decide the parameters of fair use regarding the unauthorized taking of protected content for AI training. This case-by-case approach will provide case-by-case answers, depending on the circumstances, the judge in question and even the circuit within the US federal justice system in which the case is heard. A legal approach inevitably leads to risk and uncertainty. That is not to say that in the end, the US will not resort to legislation. Thee US Justice Department recently submitted a “Statement of Interest” in the OpenAI v New York Times case, stating that a ruling in favour of the Times and other news industry plaintiffs would “threaten US national security, give a competitive advantage to foreign adversaries”, and thwart “creative and scientific progress while hindering American prosperity”. The Statement of Interest tries to put its thumb on the scale of justice, weighing in in support of the OpenAI argument that its use of NYT content was “transformational”, and thus constituted fair use. The Court is under no obligation to accept this interpretation; its duty is to interpret the law as it stands. But if the DOJ statement represents the Trump Administration’s interpretation of the law, and that interpretation is not upheld by the courts, that potentially opens to door to legislative change, assuming that Congress can be convinced to act in the name of big tech. Either way, there is continued uncertainty.
The EU approach is based on a narrow text and data mining exception embedded in the Copyright Directive. Resort to the exception is limited in two ways; if rightsholders cannot opt-out, the exception is limited to scientific organizations and cultural heritage institutions for research purposes only. There is also a general exception that allows for commercial use but in this case, rightsholders can reserve their rights by opting out.
Japan is considered by many to have a very AI-friendly regime in terms of allowing copying for training purposes, but it is not as wide as many have claimed, as I pointed out in this blog posting a couple of years ago. South Korea is struggling with the issue, being whipsawed between AI developers promising the moon and its vibrant cultural industry. China likewise is still trying to determine which way to go. It does not have a text and data mining (TDM) in its law, nor does it apply a fair use doctrine. However, it has announced plans to deal with AI training rules through its copyright development plan covering the period 2026-30.
The UK has had several false starts. A TDM exception limited to non-commercial purposes only has been on the books for several years, but the Labour government tried to expand this by advocating for a regime that would allow AI developers to freely use copyrighted content for algorithm training unless rightsholders opted out. After a massive outcry and pressure from British cultural industries and prominent performers, the government retreated. Now Britain has gone back to the drawing board.
In the case of Canada, everything is up in the air. Canada has a fair dealing regime with no TDM exception. It has held several public consultations and is developing an AI strategy but has been very unclear about where the government comes down on the issue of freely using copyrighted material for AI training. Meanwhile several Canadian media outlets are suing OpenAI for copyright infringement. OpenAI lost its argument that Canadian courts had no jurisdiction in this case. Now the Canadian music collective SOCAN is suing AI music company Suno. While AI companies are arguing for a wide TDM exception in Canadian law, rightsholders are seeking to retain the current law and use it to protect their rights.
Australia on the other hand has spoken out strongly in defence of its cultural sector and has declared that it will not legislate a TDM exception. Australian Prime Minister Albanese has taken that a step further declaring that, “An artist’s creative endeavour is their work and their property. No company should use Australian books, music, art or news to build or train AI without the artist’s control. That includes the artist’s control of the price and value of their work.” What will happen next in Australia is unclear although some tech platforms continue to try to push various compulsory licensing regimes.
As you can see, the rules are diffuse and in flux. Frankly, it is all over the map. One thing seems to be clear. The push by the tech industry for widespread TDM exceptions seems to be stalling. The litigation route so favoured in the US is yielding uncertain results. Legal actions in fair dealing countries like Canada, the UK and Australia may bring varying results with complex rules governing applicability of judgements. All this leads to greater uncertainty, and costs. However, as mentioned above, there is one universal remedy, one that transcends all the jurisdictional issues and conflicting interpretations. It is called voluntary licensing.
Voluntary licensing between rightsholders and AI developers eliminates the jurisdictional problems, assuming the agreements are structured to provide wide coverage and indemnity. With a voluntary licence, it doesn’t matter if there is or isn’t a text and data mining exception, or if the use is for commercial purposes, or if the use is sufficiently transformational. It doesn’t matter whether the AI developer is truly “enjoying” (accessing the essence of the content), as in the case of Japan. Some jurisdictions have toyed with the idea of a compulsory licence, as in India, but not only has this been opposed by both rightsholders and the AI industry, it is jurisdictionally bound. Voluntary licences deal with this issue and, if properly structured, remove the hazard of conflicting jurisdictional rulings.
While it is not always easy to reach such deals, as the OpenAI v New York Times case illustrates, it is surely better and less costly than the alternative of endless litigation. This is why it is becoming the preferred solution. More and more licensing agreements are being negotiated across the full range of copyrighted content, including audio-visual, music and publishing industries for AI training and use. Voluntary licensing is a global solution–but it requires AI companies to recognize the value of the content they want for training, and to accept the right of creators to control the use of their content. If it requires a willing buyer, it also requires a willing seller and while not all rightsholders will be willing to sign a licensing agreement, the holdouts will be the exceptions if fair deals are offered. It is a far better solution for both sides, AI developers and rightsholders, than endless litigation or constant lobbying for (or pushing back against) legislative change. It’s a global solution to a global problem.
[i] Any exception must satisfy three requirements: (1) Limited to certain special cases:. i.e. the exception must be narrowly defined and clearly circumscribed, rather than acting as a broad or general exemption; (2) No conflict with normal exploitation: i.e.the use must not interfere with the ways the copyright owner routinely makes money or manages the market for their work; (3) No unreasonable prejudice to legitimate interests of rightsholders, i.e. the use must not cause unfair or excessive economic or legal harm to the author or right-holder
Australian Prime Minister Anthony Albanese caught global attention in mid-July with his speech at the University of Sydney, “AI in Australia’s Interests”. That address, which outlined the Australian government’s approach to AI, included important statements regarding the work of Australia’s creative community, which Albanese declared is “not up for grabs”. Having previously ruled out the creation of a copyright loophole for AI training, known in the trade as a “Text and Data Mining” (TDM) exception, Albanese went on to say that;
“Australian writers, musicians, artists and journalists must retain ownership and control of their work…No company should use Australian books, music, art or news to build or train AI without the artist’s control…of the price and value of their work. Anything less is theft.”
Bravo! Hopefully this means what it says, that rightsholders will retain control and receive compensation on their terms if their works are used for AI training, even though there is still strong pressure from segments of the AI world for Australia to loosen its terms of copyright protection. Anthropic is dangling a $20 billion carrot in the form of potential investment in AI data centres in Australia but only if there is a copyright carve-out for AI training. One Anthropic proposal was for the creation of a $350 million fund to compensate rightsholders, paid for by the AI industry, but it was unclear how this would operate or if rightsholders would be required to opt-out if they didn’t want their content used. For creators to be able to control their work within the accepted framework of copyright law, they must have the ability to accept or reject voluntary licensing and not be required to opt-out of a compulsory scheme where what they receive is decided by bureaucrats on the basis of limited contributions to a common fund. The commitments that Albanese made in his speech indicate that voluntary licensing is the preferred solution, but the issue remains under study (hopefully this will be done with full transparency, following suggestions earlier this year of secret dealings on this critical matter of public interest). Nonetheless, the bright line that Albanese has drawn with regard to creatives retaining control of their work is welcome and will hopefully encourage other governments who are developing AI policy to do the same.
As noted, the AI industry never gives up. Having been clearly told that creators should retain control over how their work is used, and be paid for that use, some companies are floating new objections based on the supposed “long tail” argument. “Long tail” is the term being used to describe the many small rightsholders who create and therefore (in theory but maybe not in practice) control use of their content, as opposed to larger rightsholders such as major publishers of books, journals and newspapers, music labels, film studios and so on with whom it is easier to engage in licensing discussions. The argument is that, even if the AI industry wanted to license the content that it uses in training (by no means a given with every company), it can’t possibly deal with the myriad of small rightsholders. Elsewhere, some AI developers have dealt with this problem by ignoring it, simply helping themselves to whatever content they wanted, without licence. In some cases, they even used pirate libraries as sources, almost daring rightsholders to bring legal challenges. Anthropic is a high-profile culprit that got caught doing this.
One idea currently being promoted by Anthropic is to create a statutory licence that would be restricted just to Australian rightsholders, with funds directed only to them. It is hard to see how this is a viable solution. Not only would it violate international norms that Australia has committed to by discriminating against foreign rightsholders in the Australian market, if it is applied universally, it would lead to a stream of royalties flowing out of Australia while doing little to help local creators. Elsewhere, various non-statutory collective licensing schemes have been proposed to deal with the issue of small rightsholders, although to date none have emerged that provide true one-stop shopping for AI developers. But a compulsory licence cannot be the solution. These large companies with stratospheric valuations (especially those getting ready to launch an IPO!) surely have the wherewithal to find ways to license the content they use and will continue to use as they develop and refine their AI models.
For the past several years the tech community has followed the “better to ask for forgiveness after rather than permission before” approach when appropriating copyright protected content to develop AI platforms. The inevitable result has been a spate of lawsuits, most of which are still ongoing, with results varying from case to case and jurisdiction to jurisdiction. A second front opened by some AI platforms has been to attempt to weaken copyright protection through legislation, focussing particularly on the introduction of a broad TDM exception. As a result of this push, many countries began studying whether to bring in a TDM carve-out, or if they had a narrow TDM exception (e.g. for non-commercial research purposes) whether to widen it. From the perspective of rightsholders, it began to look like a race to the bottom. Now, the tide seems to be changing. Albanese’s speech is a good example of the shift, but Australia is not alone in taking a more considered and balanced approach.
Hong Kong is a case in point. Hong Kong has always had one of the stronger intellectual property regimes in the region, a source of competitive advantage. But under the pressure of blandishments from the tech community, it issued a consultation paper in 2024 that, among other things, proposed a TDM exception–including for commercial purposes. When the paper went out for public comment, the predictable suspects urged adoption of a TDM provision while rightsholder groups argued that existing law and voluntary licensing was suitably flexible to handle issues arising from new technologies. Now the Hong Kong Government has concluded there is no pressing need to pursue a TDM exception through legislation. The priority will be to issue best-practice guidelines grounded in the existing legal framework, providing reference for stakeholders on copyright protection and infringement liability relating to AI-generated works. Voluntary licensing arrangements between copyright owners and AI developers are seen as the answer.
I noted in a blog post last year that the TDM issue was under review in a number of Asian jurisdictions, including India, Malaysia and Korea. In India, a government-commissioned working group recommended a compulsory licence scheme that displeased just about everyone, from the tech gurus to India’s cultural creators, although to the working group’s credit, it also rejected a TDM exception for India. The lack of momentum for this idea means that current copyright legislation continues to apply, albeit with Indian courts starting to weigh in, (but unfortunately in ways that could potentially undermine India’s fair dealing balance). Malaysia continues to attract significant investment from the hi-tech sector despite not yet adopting a TDM exception, in stark contrast to neighbouring Singapore. In Thailand, which has one of the richest cultural traditions in Asia through its audiovisual and music sectors, the tech sector has been agitating for introduction of a TDM, as elsewhere. This would put at risk a content sector that is a significant contributor to national GDP as well as driver of the all-important tourist industry.
Korea is another country where the AI industry is bringing pressure to bear. Korea’s unique and thriving culture is one of its national assets, along with a domestic innovation and hi-tech industry that is second to none. To allow offshore tech companies to plunder Korea’s rich cultural tradition would be a shortsighted and shameful sellout. But neither Korea, nor Thailand, nor Malaysia, India or Australia are opposed to the development of AI, and all want their share of the benefits of this emerging if not already emergent industry. All recognize that AI development requires vast amounts of content, including the kind of high-quality content produced by their respective creative sectors.
There is a proven path for the AI industry to access this rich content. It is called voluntary licensing, and it already exists within the established four corners of copyright law. Moreover, it is increasingly becoming the solution as AI developers finally come to realize (a) that it is not worth risking the viability of the company on an unpredictable lawsuit that could result in crippling damages (or as in the case of India, years of court trials) and (b) equally important, their global push to get a legislated copyright loophole through TDM provisions in national law is going nowhere. In fact, the TDM tide is receding, as the examples of Australia and Hong Kong clearly show.
As the TDM tide goes out, the voluntary licensing tide flows in, floating the boats of both the creative and hi-tech sectors. Australia hoisted the first signal, but other jurisdictions in Asia seem ready to follow.
In early June Canada issued its national AI strategy paper, “AI for All”. As I noted in a blog post at the time, while the strategy covered many elements of AI in its 50 pages outlining policy objectives and planned actions, it managed to avoid using the word “copyright” even once. Australia has just come out with its own updated AI policy statement “AI in Australia’s interest”, which builds on its own “National AI Plan”, released last December. But whereas the Carney government in its AI strategy managed to completely avoid putting copyright into the AI equation, Prime Minister Albanese, after discussing the importance of developing AI for Australia, had this to say;
“But let me make this crystal clear: not everything produced in Australia is up for grabs.
Not at all.
Australian writers, musicians, artists and journalists must retain ownership and control of their work.
Our laws will spell that out, plain as day.
An artist’s creative endeavour is their work and their property.
No company should use Australian books, music, art or news to build or train AI without the artist’s control.
That includes the artist’s control of the price and value of their work.
Anything less, is theft.”
Blunt, clear and refreshing. If Australia can protect its cultural community while promoting policies for sensible AI adoption and development, then so can Canada.
Both Canada and Australia currently have no Text and Data Mining (TDM) exception in their copyright law. This legal loophole would allow AI developers to appropriate content without permission for training purposes. In both countries there have been calls from the tech community to introduce a TDM exception, a carte blanche that would allow AI companies to ingest copyrighted content without authorization, payment or even acknowledgement. In its December “National AI Plan”, which is much more analogous to Canada’s “AI for All” than Albanese’s recent short AI policy statement–in that it outlined a range of detailed policy proposals for AI adoption in Australia– the Australian government nonetheless managed to grasp the copyright nettle unambiguously.
Among the issues highlighted under “AI Risks and Harms” was the following:
“Reviewing application of copyright law in AI contexts: The Attorney-General’s Department is engaging with stakeholders through the Copyright and AI Reference Group to consult on possible updates to Australia’s copyright laws as they relate to AI. The government has provided certainty to Australian creators andmedia workers by ruling out a text and data mining exception in Australian copyright law”(emphasis added)
Just as the Australian government has sensibly ruled out a TDM option. Canada needs to do the same, as called for Canadian cultural umbrella groups, such as the Coalition for Diversity of Cultural Expression (CDCE).
So far Canada has danced around the issue. Heritage and Identity Minister Marc Miller has said that “the current copyright law does and should protect those that have created material, and people need to be compensated properly”, but he is just one minister among several. Evan Solomon, Minister of Artificial Intelligence and Digital Innovation, and Minister of Industry Melanie Joly, both have a big piece of this file. One can expect that both can be counted on to be more sympathetic to tech bros than cultural mavens. What is needed is a prime ministerial pronouncement clarifying that Canada’s creative community–artists, writers, publishers, musicians, filmmakers, photographers, journalists and more– is not going to be thrown under the bus on the pretence of keeping Canada competitive in the global AI game.
In the wake of Australia’s announcement that a TDM exception was off the table, the tech industry tried a new approach by suggesting the creation of a centralized fund that would be used to compensate rightsholders for the permissionless use of their works in AI training. Specifically, AI company Anthropic reportedly tied a proposed $15 billion USD ($21.6 billion AUD) investment in data centres in Australia to creation of the creatives fund in order to allow to access Australian content without licensing or negotiation with rightsholders. Australia’s creative community quickly mobilized. Their concerns were heard. Along with setting clear guardrails ruling out the unauthorized use of copyrighted creative works, Albanese has created a new Office of AI within the Prime Minister’s Office, recognizing the need for policy coordination given the breadth of AI’s policy impact. This is something that Canada might consider. It has Evan Solomon, Minister of Artificial Intelligence and Digital Innovation, but there seem to be very few cultural community voices within Solomon’s hearing range.
Australia has the same goal as Canada of getting its fair share of the AI pie while managing AI adoption and its impact on society. But there is one big difference. In so doing, the Australian government has made it clear it will pursue its AI goals while simultaneously respecting and protecting its culture and its creators. Canada’s cultural and creative community deserves no less consideration.
There is an ongoing struggle between the tech world of AI training and the cultural world of content creation. It has led to lots of litigation but also an increasing number of licensing agreements, the obvious market solution. Litigation has helped convince AI companies to share some of the wealth by pursuing licensing. Yet the AI world continues to try to find ways to avoid the basic step of seeking permission from rightsholders for using their valuable content to create their products.
Anyone who has seen the striking graphic “Who is Suing Whom in AI”, created by the design website Information is Beautiful, will be struck by the enormity and breadth of the issue which is so cleverly displayed, with the big AI developers such as Perplexity, Anthropic, Meta, Google, Open AI, Midjourney, Cohere and others at the centre with the creators (every content entity from Conde Nast, Getty Images, Universal Music Group, CNN, Disney and Thomson Reuters to Elsevier, Dow Jones, New York Times and others) ranged around the periphery, a stunning visual encompassing more than 100 lawsuits in the United States. That graphic was up-to-date as of June 26 of this year. Since then, at least one more major lawsuit has been filed, by a group of textbook authors against Meta. The graphic does not include the first such case in Canada where a group of media organizations (Canadian Press, Torstar, The Globe and Mail, Postmedia and CBC/Radio-Canada) is suing OpenAI, or the Getty Images case in the UK, or indeed any cases outside the US. From this graphic, it would seem that to resolve the issue of how copyrighted content is going to be used in AI development and training, litigation is the inevitable route. But is it?
As far as I am aware, Information is Beautiful has not created a similar graphic to display the range of licensing deals that have taken place, many of them between some of the same actors that appear on the litigation chart. If they did it would be similar, but encompassing even more licensing agreements than lawsuits. Licensing deals are being struck so frequently it is just as hard to keep up with them as it is to track all the litigation underway. The University of Glasgow’s CREATe Centre says it has documented 274 licensing deals and has a chart that tracks 109 of them. Whatever the number, it is a lot and it is growing. That is not to say that the AI industry has finally accepted the need to pay for the content they are using to create their products, just as they pay for software engineers or data processing capacity. This is where the link between litigation and licensing becomes interesting.
In a perfect world, AI developers would obtain their inputs through the market on the basis of permission, which would encompass both compensation (in most cases) plus transparency or accountability, i.e. documenting what content was used. But we don’t live in a perfect world, which is why we have the rule of law and courts to enforce those laws. In some cases, AI platforms did begin negotiations with rightsholders but when it was not possible to reach an agreement, the AI industry switched tactics and took the content anyway, arguing it was legal to do so for a variety of reasons. This is precisely the scenario that led to the New York Times suing OpenAI. These cases are even more egregious because there was initially a tacit acknowledgement by the user that the content had value. Then, when the price or conditions did not suit the potential licencee, suddenly it was okay to take the content anyway under the guise of fair use. Various arguments have been deployed ranging from the claim that no copying actually occurs, to the dubious assertion that what is copied is data not content, to the invocation of the US “transformation” doctrine.
On the issue of copying, a study by the Atlantic (AI’s Memorization Crisis: Large language models don’t “learn”—they copy. And that could change everything for the tech industry) convincingly demonstrated the uncomfortable truth that LLMs can reproduce long excerpts from books they have been trained on. The inputs are not just ones and zeros, they are content— someone else’s content that was taken without permission. Whether the use was fair according to US fair use interpretations is still an open question. US courts and other countries are trying to come to grips with this issue. In countries such as Canada or Australia, where there is no statutory copyright exception for Text and Data Mining (TDM) that would permit permissionless AI training on content, the AI industry has been floating various workaround proposals. The “incentives” would include (in Australia) establishing a government-managed fund to compensate rightsholders according to some sort of formula, plus investments in AI data centres. What is missing from proposals such as this is the concept of permission from those who actually own the content, or even discussion of the proposal with them. As Prof. Rod Sims, former Chair of Australian Competition and Consumer Commission, put it in a recent opinion piece in Canada’s National Post, “what other sector refuses to negotiate with suppliers and instead goes to government to bypass such a step?”
Let me use a food industry analogy to make the point even more clearly. When you run a restaurant you have labour costs, rent, taxes, etc. and the cost of ingredients to consider. You don’t get to raid the farmer’s field to obtain your inputs for free, just because you are able to root out crops without the farmer being able to stop you or even know it is happening. Setting up a fund to “compensate” farmers for their stolen crops, on terms set by the government rather than the market, doesn’t even begin to make this right. Legalization of this theft would remove any possibility of litigation or legal protection, for the farmer—or for content owners. Litigation, while protracted, costly and potentially leading to uncertain outcomes, is nonetheless the stick that is needed to facilitate licensing.
The obvious route for the AI industry to take is to license the content they want to use. That may not seem as “efficient” as just taking it for free but with the threat of litigation hanging over the proceedings, licensing suddenly becomes the more efficient alternative. It is also win/win for both AI developers and the content industries. And, it is simply the “right thing to do”.
I am pleased to note that this blog was recognized by Feedspot as being among the “40 Best Copyright Blogs to Follow in 2026”. In fact, we hit the middle of the pack at No. 20. I am honoured to be included in such distinguished company.
Feedspot is an RSS Reader that lets readers subscribe to blogs, news sites, and any website they wish to follow.
New Zealand is proposing to introduce a number of optional updates to its Copyright Act when it enacts required changes to bring legislation into compliance with two treaties it has signed. This is good news for creators. Still to be addressed, however, is the thorny issue of AI training on copyrighted content.
New Zealand needs to make some required legislative changes to its Copyright ordinance as part of implementing two treaties it has signed, the UK-New Zealand Free Trade Agreement (FTA) and New Zealand’s FTA with the European Union. In both cases New Zealand has agreed to extend its term of copyright protection from life of the author plus 50 years to “life plus 70”, as well as preventing the circumvention of TPMs (technical protection measures, aka “digital locks”) except in specified narrow situations. These provisions must be enacted by May of 2028. They will bring New Zealand’s copyright law into alignment with most of its major trading partners. However, while there is a legal requirement to address the above two issues, the Ministry of Business, Innovation and Employment (MBIE) has proposed that a number of other copyright issues also be addressed as part of the process of updating the Act. These include;
supporting not-for-profit gallery, library, archive and museum (GLAM) organisations to preserve and provide access to collections, including by allowing use of orphan works, making digital copies for preservation and access, and applying research and private study copying rules across all GLAM organisations, with safeguards for copyright owners
introducing a new fair dealing exception for parody and satire, applying across a wide range of works while maintaining authors’ moral rights
providing courts with a framework to order internet service providers to block access to overseas websites primarily engaged in copyright infringement, with appropriate safeguards and flexibility
removing an outdated peer-to-peer file-sharing enforcement regime that is no longer used, reducing compliance costs for internet service providers
enabling copyright licensing organisations to take collective action on behalf of copyright owners to prevent infringement
clarifying that the first distribution right is only exhausted where the copyright owner has consented to the overseas sale of copies, supporting control over parallel imports of infringing copies
changing the default rule for commissioned works so that creators are the first copyright owners unless agreed otherwise
extending resale royalty rights for visual artists by 20 years to align with the longer copyright term.
It is encouraging to see New Zealand take this opportunity to review and update its copyright framework while it implements the needed changes to meet its trade agreement commitments. Canada was also required to extend its copyright term as a result of the new NAFTA agreement with the United States, and it did so, at the last minute. However, it did the minimum required and passed on the opportunity to address wider issues, of which many have been identified by Parliamentary committees, while more are coming forward as a result of developments in AI.
The proposed changes in New Zealand should be welcomed by the copyright and copyright-using community. They will provide legal protection for the sort of digital replication that the GLAM sector needs to preserve older and orphan works, although more information on what how the research and private copy rules will be implemented is needed. Widening fair dealing to include satire and parody has been done in a number of jurisdictions, and this will bring New Zealand in line with other Commonwealth countries like Australia, Canada and the UK that have such exceptions (“parody, caricature, and pastiche” in the wording of the UK legislation). In the application of the defence, New Zealand courts should follow the Australian lead, where courts have kept a tight rein on this defence. Parody is a tricky exception to invoke, as a recent UK case well illustrates. The moral rights of the author are also a factor to consider.
For the first time, site-blocking (that is, requiring ISPs to block pirate offshore websites, after legal review) will have a firm foundation in New Zealand law. Australia has had such legislation on the books for more than a decade, and the UK for longer than that. Both the UK and EU treaties required New Zealand to allow the courts to issue injunctions “against an intermediary whose services are used by a third party to infringe intellectual property rights.” Canada has dealt with this issue through the courts exercising their inherent jurisdiction without the enactment of specific site-blocking legislation, with initial challenges from some ISPs being dismissed on appeal. The process has now become routine. It seems the New Zealand government intends to ensure clarity by amending copyright legislation to “provide courts with a framework to order internet service providers to block access to overseas websites”. IP scholars in New Zealand, such as Prof. Graeme Austin, have been calling for the government to take the lead. It seems they have been heard.
The empowering of collective management organizations (CMOs) to take legal action against infringers on behalf of their members is also an important step. Under present provisions, CMOs cannot bring actions because they do not hold the rights to individual works. This requires multiple authors either to take individual actions or join in a joint action. Given the cost of such an exercise, this is not feasible (large publishers who have licensed rights from authors may be in a position to do this, but authors themselves are hamstrung). Giving their collective management organization the right to represent them is a positive move. This is a move that Canada could well replicate to enable CMOs like Access Copyright to represent authors.
Changing the default rule for commissioned works will, for example, give photographers greater control over their work. Clients can contract for the right to display copies of the work but the copyright in the original work will belong to the creator. The same is true for artistic works unless there is a specific agreement that the work is created under an employment contract. Canada enacted this provision in 2012 when it passed the Copyright Modernization Act. Extending the resale royalty rights for authors to match the longer copyright term keeps these two provisions in alignment. New Zealand, like Australia and the UK, and EU member states, has enacted an Artists’ Resale Right (ARR), which allows a small portion of the proceeds of a resale of artwork through a professional dealer to be paid to the original artist (or their estate). Canada has been promising for several years to enact an ARR but has not yet done so.
The one big issue this round of copyright amendments will not address is use of copyrighted content for AI training. That is a rapidly evolving issue in many countries and is a moving target. The solution, as suggested in this article by Prof. Austin, is to foster market solutions, that is facilitating the licensing of content to AI developers. The way not to do this is to provide a wide exemption for AI training, as many in the tech world are advocating, but to ensure that rightsholders have the right to protect their content and to grant access to it on terms that they agree to. This is already happening in a number of areas such as licensing agreements between major publishers, news enterprises, and the AI industry, but individual authors are still being left out of the discussions.
Australia has just ruled out creating a fair dealing exemption for AI training (known as the TDM or Text and Data Mining exemption). Even the notoriously anti-copyright Productivity Commission supports this position. Such an exemption would remove any incentive for AI developers to negotiate with rightsholders for use of content. Hopefully New Zealand will follow suit in this regard. While we will have to wait for further developments when it comes to dealing with AI issues, the current set of proposals will be very useful in renewing and updating the copyright framework in New Zealand.
We read daily about new lawsuits brought by rightsholders against AI developers, strategy papers floated by governments seeking to solve the riddle of reconciling copyright and AI, and declarations issued by authors proclaiming the end of human creativity. The creative community seems to have coalesced around the principles of transparency, permission and remuneration but the tools to effect those key elements remain elusive. The AI community would generally prefer not to pay or ask permission but is gradually accepting the need to license content. Yet there is still a technical gap in terms of knowing what content has been used, when and how. Without that knowledge, the principles of permission and remuneration are left treading water. The blog post below by Jim Bryant, Co-Founder and CEO, Citations LLC, offers potential solutions to this challenge, and I offer it to you as a possible pathway forward. I have no financial interest in Citations LLC, nor did they pay me to post this information. It is presented as a contribution to the search for a world where copyright and AI can co-exist for mutual benefit. (Hugh Stephens)
A problem or an opportunity?
Imagine a student in Montreal asks an AI assistant a question about traditional Chinese medicine, in French. The AI answers fluently — in French — drawing on the Encyclopedia of China, a monumental work with over 125 million characters that has never been translated into any language. Now imagine the same student switches to English and asks a follow-up question. The AI answers again, equally fluently, in English. The student is satisfied. The publisher gets nothing. No notification, no attribution, no compensation. They don’t even know it happened.
This scenario is entirely plausible with current AI technology. And while it represents a genuine copyright problem — real-time AI translation of a protected work, without license, in a jurisdiction whose law was not written to contemplate it — it also represents something else: an extraordinary, unrealized opportunity.
For the first time in the history of publishing, the technology exists to know, at the moment it happens, that someone in Montreal, Mumbai, or Mexico City is asking a question that your content just answered. The question is whether publishers will help build the systems to capture that signal — or whether they will leave it entirely to the AI companies, who are already building without them.
Publishers have always been flying blind.
Think about what publishers have never been able to know — and what AI companies, for the first time, can. An AI system that has trained on your works without permission is, in effect, drawing on your content every time it answers a relevant question. Which of your titles is it using right now, and where? Which readers are getting answers derived from your content without ever being directed back to the original? Which backlist titles are generating AI responses in markets where you have no distribution and no visibility? Which gaps in your catalogue are readers repeatedly trying to fill — and how would you know, if the only signal is buried inside a system you have no access to? The argument for independent monitoring infrastructure is not only about compensation. It is about visibility. Publishers are currently funding AI responses with their content and receiving nothing in return — not money, not data, not even the knowledge that it is happening.
For centuries, publishers sent their works into the world and largely lost sight of them. Sales data arrived months or years later, filtered through agents, booksellers, distributors, and described what sold — not what readers wanted but couldn’t find. The feedback loop from reader demand to editorial decision has always been slow, indirect, and incomplete.
A properly instrumented knowledge access infrastructure changes all of that. Real-time query data across AI systems is, in effect, a continuous signal of what readers want — more granular, more current, and more honest than any market research tool the industry has ever had. That data is a byproduct of the same system that creates the copyright exposure publishers are currently fighting in court.
The moment of demand is the moment to act.
Here is the specific opportunity that AI creates, and that no prior technology has made possible: when an AI system surfaces content in response to a query, it creates a demonstrated moment of demand. A reader who just received an AI-generated answer drawn from a specific book is, at that moment, maximally interested in that book. That is the moment to offer them the chance to borrow it from a library, purchase it from a retailer, or access an authorized digital edition.
Rather than substituting for the book, the AI interaction becomes the discovery mechanism that leads to it. Publishers have spent decades trying to close the distance between the moment a reader becomes interested in a title and the moment they act on that interest. AI closes that distance to zero — but only if the infrastructure exists to capture it. Without that infrastructure, the moment passes, the reader moves on, and the publisher never knew the opportunity existed.
Libraries are being bypassed — and publishers are losing their best customers. Libraries are among the largest single customers some publishers have. A major academic or reference publisher may depend on library subscriptions for a substantial share of its revenue. AI is disrupting that relationship in ways that have received too little attention. When a patron who would previously have borrowed a book — or prompted their library to acquire it — instead receives an AI-generated answer derived from that same book, the library never makes the purchase, the publisher never sees the revenue, and neither institution knows the transaction occurred. The AI company captures the value; the library loses a use case; the publisher loses a sale. The institution most structurally committed to legal, compensated access to knowledge is being systematically bypassed by systems that obtained that knowledge without payment.
The same logic applies to real-time trend identification. Aggregate query patterns across an AI knowledge system are a leading indicator of what readers want — not what they bought last quarter, but what they are looking for right now. Which subjects are rising? Which titles are being asked about in markets where they have no distribution? Which authors are generating interest that isn’t yet reflected in sales? This intelligence, continuously available, would transform publishing from a reactive industry into a responsive one.
The translation question is the hardest — and the most important.
The Encyclopedia of China example is worth dwelling on, because it illustrates both the opportunity and the complexity in their sharpest form. That encyclopedia has never been translated — into French, English, or any other language. The economics of translation have made it prohibitive: 125 million characters, uncertain commercial return, no obvious path to a global audience. As a result, it has been accessible only to readers of Chinese. That constraint has nothing to do with the quality or the importance of the content.
AI removes that constraint entirely. In this hypothetical, a reader anywhere in the world could ask the encyclopedia a question in their own language and receive an answer. This is, genuinely, one of the most remarkable things that AI makes possible: the dissolution of language as a barrier to knowledge, overnight, at no marginal cost.
But it raises a set of copyright questions that existing law is not equipped to answer. A real-time AI translation is, in the most precise legal sense, the creation of a derivative work — at the point of query, in a foreign jurisdiction, without a license, without attribution, and without compensation to the original publisher. It is not covered by any existing text-and-data-mining exception, because it is not mining — it is real-time derivation. It is not covered by fair use or fair dealing analysis that was designed for static reproduction, not dynamic on-the-fly translation.
And yet the underlying interest of the publisher is not to prevent this from happening — it is to be compensated when it does, and to have some say in how their content is represented. A framework that would allow the publisher of the Encyclopedia of China to authorize AI-mediated translation under defined conditions, receive a per-query payment, and have the source attributed, would serve everyone’s interests. The absence of such a framework means the publisher gets nothing, the AI company gets everything, and the reader gets an answer of uncertain provenance.
The ten copyright challenges — briefly.
It is worth cataloguing the specific challenges, because they are often discussed in isolation when they actually share a common cause. The publishing industry currently faces at least ten major copyright issues arising from AI:
AI training — whether training on copyrighted works requires permission and compensation, currently being litigated in multiple jurisdictions.
Transparency — AI developers do not disclose what content their models were trained on, making it impossible for rights holders to assess exposure or negotiate terms.
Reproduction — models can and do reproduce passages that closely approximate protected expression, as documented in peer-reviewed computer science research.
Market impact — AI summaries and Q&A responses can substitute for the original work, displacing revenues that would otherwise flow to the publisher.
Derivative works — the degree of transformation required to render AI output non-infringing remains genuinely unsettled, particularly for outputs that blend multiple protected sources.
Attribution — AI outputs routinely fail to identify the works they draw on, undermining both the moral rights of authors and the practical basis for any royalty mechanism.
Compensation — no industry standard governs AI licensing fees; per-query, per-token, and blanket models are all being proposed, with no settled framework.
Retrieval — retrieval-augmented generation systems access copyrighted content at inference time, raising rights questions distinct from and additional to those arising from training.
International law — training data crosses borders; copyright law does not; EU, US, UK, Japanese, and Canadian frameworks diverge in ways that create genuine compliance complexity.
Auditability — without verifiable records of what was accessed, when, and in what context, no licensing agreement is enforceable and no royalty calculation is credible.
These are not ten separate legal problems. They are ten symptoms of one missing piece of infrastructure: a neutral, independent system for monitoring how AI systems access and use copyrighted content, reporting on that usage in real time, and enabling settlement between AI platforms and rights holders on the basis of verified data rather than estimates.
Why the infrastructure must be independent.
This point deserves emphasis, because there is a tempting shortcut that would not actually work. Publishers cannot rely on AI developers to build and operate the systems that monitor AI’s use of their content. The conflict of interest is structural: the party whose compliance is being measured cannot be the party doing the measuring.
What is required is a neutral layer — operated independently of both AI developers and publishers — that records access events, aggregates usage data, reports to rights holders, and enables automated settlement. Think of it as the knowledge economy’s equivalent of a financial clearinghouse: not owned by any single participant, trusted by all of them, and essential to the functioning of the market.
This is not a novel concept — it is exactly the model that makes collective rights management organizations function in the music industry and payment card networks function in financial services. Every industry that has needed to account for consumption at scale and distribute revenues to multiple rights holders has eventually built a neutral clearinghouse.
The window is open — but not indefinitely.
Canada’s AI strategy, recently released, makes almost no mention of copyright or the rights of content creators — a significant omission that Hugh has written about on this blog. The European Parliament’s work on AI and copyright has moved further, but still focuses primarily on training rather than on the access and retrieval layer where the most tractable opportunities lie.
The practices governing how AI systems access knowledge are being established right now, largely by default. The companies building AI systems are not waiting for a legal or regulatory framework; they are building, and the norms are hardening around what they build. Publishers who are not at the table when that infrastructure is designed will find themselves subject to whatever framework others have built for them.
Copyright law exists to balance access and incentive — to ensure that knowledge can circulate while the conditions that make knowledge production sustainable are preserved. AI does not change that objective. It changes the technical conditions under which the balance has to be achieved. The good news is that those technical conditions, for the first time, make real-time monitoring, attribution, and settlement not just possible but straightforward.
The question is not whether AI will access books. It will. The question is whether publishers will be watching when it does — and whether they will have built the systems to act on what they see.
That system already exists. It logs the moment, attributes the source, and settles the account — not as a future framework, but as infrastructure operating today. It’s called Citations, and it’s already watching.
* * *
About the author
Jim Bryant is the co-founder and CEO of Citations LLC, which has built the independent infrastructure for rights-aware AI access to authoritative content — enabling real-time monitoring, attribution, and settlement between AI platforms and publishers. See how it works at: citationslogic.ai. Jim previously founded ProCD, one of the first CD-ROM reference publishing companies; managed Information Please, which became one of the most visited reference destinations of the early internet; and founded Trajectory, which developed and deployed natural language processing algorithms to read and extract structured metadata from over one million books in English and Chinese.
If it does, you would never know it from reading Canada’s new AI strategy just released by the Minister of Artificial Intelligence and Digital Innovation, Evan Solomon. It is a magistral document, addressing key elements of AI under six pillars: (with my shorthand summary in brackets)
Protecting Canadians and safeguarding democracy (addressing trust, safety and privacy concerns)
Empowering Canadians (promotingAI literacy and economic opportunity)
Powering AI adoption for shared prosperity (accelerating adoption, especially for SMEs)
Building a sovereign AI foundation (building domestic compute, cloud and connectivity infrastructure)
Scaling Canadian champions (more government funding for domestic AI development)
Building trusted economic and governance partnerships and global alliances (leading the creation of a multinational middle power alliance to curb the power of hegemons and hyperscalers)
The latter objective will no doubt go down really well with the Trump Administration!
Those six headings cover just about all aspects of AI, from its creation to its use to its impact on the economy, on society and on individuals. But in all 50 pages of the document, as far as I can ascertain, you won’t find the word “copyright”, although “protecting intellectual property” is certainly featured. The intellectual property rights that are mentioned have nothing to do with the rights of those whose content was used without authorization to create AI but rather relate to protecting the intellectual output of AI developers in Canada. John Degen, CEO of the The Writers’ Union of Canada (TWUC) was the first to call this out. Given the make-up of the task force that produced the report, this is not surprising. While it was made up of the great and the good from the AI world, with academics, financiers, CEOs, cybersecurity experts, innovators, educators and so on as part of the roster, there was not a single representative from the cultural community.
There are many elements of AI this document tries to address, all of them important to a country like Canada, although there are limits to what can be done by a middle power given that the lead on development has been seized by a handful of large companies, mostly in the US. The US government itself is caught in the dilemma of wanting the US to lead AI development yet not becoming overwhelmed by it to the point that a few major corporations are calling all the shots.
As for content issues, including what must surely include some copyrighted content, they are addressed only indirectly in the Canadian strategy. The three principal issues relating to content are; (1) privacy and access to data; (2) Canadian identity and culture; and (3) AI misuse, such as creation of deepfakes and misinformation.
On privacy and data, the document notes that AI is only as powerful as the data it can access (how true!). It reminds us that governments in Canada hold vast amounts of data that should be treated as a strategic national asset and mobilized to fuel innovation and productivity (i.e. provided for AI research). Thankfully, there is a tip of the hat to the need for “strong privacy protections” but there is no mention of the unauthorized scraping of databases and protected content by AI developers, both domestic and international. Privacy is important but so is ownership of content, and the right to grant permission to use it. Unfortunately, this latter point is not mentioned.
Protecting and promoting Canadian identity and culture is also mentioned as an important goal. It is obvious that if AI developers are blocked or hindered from ingesting Canadian content, then there will be less of Canada reflected in AI outputs. That argument was put forward recently by Michael Geist in a blog post criticizing recommendations issued by the Parliamentary Standing Committee on Heritage that had called for protection of the property rights and interests of artists through the Copyright Act on the basis of authorization, remuneration and transparency. This would lead to “AI without Canada”, according to Prof. Geist. This could be true if AI developers did not need or want curated Canadian content, but they do. The solution, as I pointed out, is not to give away everything in the shop window by creating a broad AI training exception in Canadian copyright law–which would amount to legalized theft, but instead to facilitate licensing solutions by resisting the smash-and-grab. Applying the existing legislation will incentivize the AI industry to strike deals with rightsholders. In other words, they will pay a negotiated amount for the products on display. That’s the best way to get more Canadian content into AI.
On the identity issue, the government’s summary document has this to say:
“Canadian AI must support, reflect, and project Canadian culture, which includes our customs, our history, and our heritage. Canadian voices, languages, communities, and knowledge must also be represented in how AI systems are designed, built, and used. Given our diverse and multicultural society, our approach to AI must acknowledge and support this rich diversity, including strengthening the French language by capturing and projecting its idioms, expressions, and cultural contexts.”
The best way to do this is to ensure that quality content in both official languages is made available to AI developers. As I have stated above, the fairest and most efficacious way to do this is through content licensing. Broad copyright exceptions will not facilitate licensing discussions. In fact, they do just the opposite by encouraging avoidance of dealing with rightsholders.
Regarding misinformation and deepfakes, this is a huge concern, and not just in Canada. Various legislative solutions have been proposed such as the bipartisan NO FAKES Act, currently working its way through the US Congress (opposed, as usual, by the internet libertarian organization, the Electronic Frontier Foundation). Other countries, such as Denmark, are addressing the issue through amendments to copyright law, giving individuals the reproduction rights to their image and voice. The UK has an anti-deepfake law on the books, introduced earlier this year, but Canada is still struggling to get its Online Harms legislation, after a couple of false starts, finalized and across the line. Re-introduction of that legislation is expected imminently, and will likely include social media restrictions on children, a highly controversial issue.
Privacy in relation to access to data, cultural identity, and misinformation including deepfakes are all content issues that Canada’s AI strategy will need to address. And so is copyright, although not mentioned in the strategy. Putting the best possible gloss on things, perhaps it is just as well there was not some throwaway line in the strategy pointing to the need to provide wider access to copyrighted content to ensure that Canada remains competitive on AI. That is the argument often employed by those who want freer access to “OPC” (Other Peoples’ Content). The argument is that “Everyone else is doing it (i.e. giving it away–which is factually untrue), so we have to as well in order to stay competitive”. Maybe silence was better than saying the wrong thing in this document.
In the absence of any reference to copyright issues, the last word must rest with Heritage and Identity Minister Marc Miller who spoke recently to the press after the National Summit on Artificial Intelligence and Culture in Banff, AB. The Minister is quoted as saying that Canadian copyright law is already clear that artists’ work needs to be respected, and that…”the current copyright law does and should protect those that have created material, and people need to be compensated properly.”
While that is encouraging, it would have been nice to have had this reaffirmed in the AI strategy document.
It was as predictable as wasps at a picnic. Within days of the Canadian Parliament’s Heritage Committee releasing its report on “The Impact of Artificial Intelligence on the Creative Industries”, with its lead recommendation being (my highlights)…
That the Government of Canada protect the property rights and interests of artists through the principles of the Copyright Act, in accordance with the ART principle—authorization, remuneration and transparency:
a) The Government of Canada must take the necessary steps and ensure that the scope of the Copyright Act applies to AI-generated content in order to guarantee copyright protection.
b) The Government of Canada must mandate greater transparency from AI developers regarding copyrighted works used to train their models, including disclosure of training data sources, to enable proper authorization and licensing.
c) The Government of Canada must establish a clear opt-in consent requirement for the use of copyrighted works in the training of artificial intelligence systems, ensuring that creators’ works may not be used for text and data mining or model development without their prior authorization.
…prolific tech and copyright commentator Michael Geist of the University of Ottawa was attacking its conclusions, issuing warnings that unless the tech industry is allowed (without authorization or compensation from rightsholders) to help itself to copyrighted content for the purpose of AI training, we will have “AI without Canada”. In other words, unless the tech industry is allowed to plunder Canadian content in the same way that it has been doing to date in the US (although this is meeting legal challenges and is quickly changing as licensing solutions take hold), there will be less Canadian content in the training data. This, apparently, will leave Canada as an “outlier” compared to peer jurisdictions. The AI developers will turn their back on Canada and rush off elsewhere. (This is a standard threat deployed by the AI industry to play off one country against another). He cites the EU, Japan, Singapore and Israel, as well as the US in support of this interpretation. Not mentioned as “peer jurisdictions” are the UK and Australia but then that would not have served the purpose of his narrative. Australia has recently declared it will not be legislating a Text and Data Mining (TDM) exception to its copyright laws to legalize unauthorized ingestion of copyrighted works for AI training, while the UK has just hit the pause button on a series of ill thought-out and badly received proposals to allow AI developers to freely use copyrighted content to train their AI algorithms unless rightsholders specifically opt out.
Singapore and Israel are among a small minority of countries that, under US pressure, have adopted US-style fair use laws that potentially allow for a weakening of copyright protection through a hodge-podge of court rulings. While many cite Japan as a jurisdiction that has given carte blanche to tech interests and AI developers, the facts are quite different as I pointed out in this blog post a couple of years ago. Japan has a strong cultural industry that it wants to nourish and protect and has defined its TDM exception very narrowly and carefully. The EU, has two provisions in its Copyright Directive related to AI training (Article 3 which permits TDM carried out only for non-commercial scientific research purposes, and Article 4, which permits TDM for any purpose, including commercial, as long as rightsholders have not opted-out, subject to strict transparency provisions by AI companies). Both impose constraints on AI developers, although there are differing views on opt-out.
Opting-out may sound like a compromise that both rightsholders and the AI industry could support but Britain’s example demonstrates otherwise. In its now aborted public consultation, the UK government put forward several options including its “preferred” option of opt-out. Fully 97 percent of respondents, from both the tech and creative communities, trashed this option. For creators, opting out not only stands copyright on its head (it is a property right, so why should holders of that right be required to notify someone who wants to infringe on that right that they may not do so, i.e. it’s like passing a law allowing anyone to picnic on my front lawn unless I post a “No Trespassing” sign), but it is technically difficult to do, especially for individuals and small-scale rightsholders. The robots.txt protocol is not binding and is in many cases not very effective. The tech industry doesn’t like opt-out because it imposes constraints on their untrammelled ability to access anyone’s copyright-protected content, anywhere, anytime. Instead the Committee recommends “a clear opt-in consent requirement” for the use of copyrighted works in the training of artificial intelligence systems.
Now it’s my turn to quibble. IMHO, there should be no explicit need for a rightsholder to “opt in”. I think that Canada’s copyright laws, properly interpreted, already provide sufficient protection to prevent unauthorized use. A rightsholder can “opt in” to AI training or any other unauthorized use not subject to fair dealing by granting a license to use their content. If that is an “opt-in” requirement then I am in favour. If yet another opt-in step is required, this would seem to be unnecessary. Licensing is a growing phenomenon. AI developers want reliable, curated content to develop their applications. As long as they are prevented from simply helping themselves, there is incentive for them to reach licensing deals with content owners. However, giving the tech industry a pass by allowing themselves to take for free whatever they want in the name of developing AI applications (for their commercial advantage) removes the needed incentive to negotiate with rightsholders. As to whether unauthorized use for AI training constitutes fair dealing, as Dr. Geist claims (“most TDM for AI training purposes would likely qualify as fair dealing under existing law”), this is doubtful to say the least. It is hard to imagine which fair dealing purpose currently applicable in Canadian law (research, private study, education, parody or satire, criticism or review, news reporting) would apply particularly when there are fair dealing limits to the amount of a work that can be used for such purposes, and specific factors that must be applied as to the effect of the dealing on the work.
The Committee’s lead recommendation is not the only complaint that Dr. Geist has about the Committee’s report. He feels it is unbalanced because the majority of its witnesses represented the cultural industries. It’s true that its lead recommendation is very much in line with the mainstream views of the Canadian cultural community. It was, after all, the Report of the Standing Committee on Canadian Heritage. This reminds me of the conflicting reports on copyright issued a few years ago by the Heritage Committee and its counterpart the INDU Committee. The 2019 Heritage Committee report, titled Shifting Paradigms, was attacked at the time by Dr. Geist as “the most one-sided Canadian copyright report issued in the past 15 years”. He claimed that there was “no attempt to engage with a broad range of stakeholders”, even though he himself appeared along with a number of others who shared his perspective on copyright. Shortly after issuing its own report, the INDU committee then issued a tone-deaf “We’re in charge” press release reminding the world that it had “sole responsibility” for administering the Copyright Act. (This is not strictly accurate). Dr. Geist’s main complaint, whether with “Shifting Paradigms” in 2019 or the current Heritage Committee report seems to be that the Committee members, in their wisdom, did not take his expert advice.
What is the function of Parliamentary Committees? It is to hear evidence, draw conclusions and make recommendations. He complains that while there were different points of view, including notably his, on how to tackle the issue under study, the Committee’s conclusions did not reflect these views. Was it because, numerically, there were more pro-copyright witnesses from the creative community that those from the Geist camp? That is theoretically possible if it were just a mathematical exercise of adding up comments in a pro and con column. But that is not the case. While the Report made a conscientious effort to capture the full range of comments, including those of Dr. Geist, in the end the members (from three political parties) made a judgement and reached consensus conclusions. (Although the Conservative Party members provided their own addendum that added to but did not refute the Committee’s conclusions). Presumably the members of the Committee were more convinced by the force of the arguments presented by some witnesses than others. Given the range and similarity of concerns presented by disparate members of the creative community it is not surprising where they came out in terms of conclusions.
Dr. Geist is entitled to disagree with these conclusions and recommendations. To be fair, his blog commentary echoes the position he presented to the Committee, except for his complaints about process. As I said at the outset, his attack on the Committee’s report is entirely predictable, like wasps at a picnic. And those wasps can be so annoying, distracting from the main event with the occasional bite and annoying buzzing, but as any determined picnic-goer knows, it’s important to not let them become the centre of attention. The Heritage Committee’s report was carefully considered and drafted by an all-party group after hearing from a wide range of experts. It provides important recommendations that the government would be well advised to take into account as it develops a legal framework in which both the AI and creative industries can co-exist and flourish.
