Canadian Heritage Committee on Oct. 8th, 2025

[The Chair, Liberal, Lisa Hepfner]

I am calling this meeting to order.

Welcome to meeting number six of the House of Commons Standing Committee on Canadian Heritage. Today's meeting is taking place in a hybrid format, pursuant to the Standing Orders.

Members are attending in person. Do we have anybody on Zoom today? No, we do not.

Please wait until I recognize you by name before speaking, and all comments should be addressed through the chair.

Pursuant to Standing Order 108(2), the committee is meeting to study the effects of the technological advances in AI on the creative industries.

It is now my pleasure to welcome our witnesses. From ArtIA, we have Marc-Olivier Ducharme, director of innovation, alliances and futures at Sporobole. From OCAD University, we have Kelly Wilhelm, head of the cultural policy hub. We have Patrick Rogers, chief executive officer at Music Canada. We have Margaret McGuffin, chief executive officer of Music Publishers Canada. From the Society of Composers, Authors and Music Publishers of Canada, otherwise known as SOCAN, we have Jennifer Brown, chief executive officer.

Welcome.

Each organization will have five minutes to speak.

We'll start with Marc‑Olivier Ducharme.

[Marc-Olivier Ducharme, Director, Innovation, Alliances and Futurs, Sporobole, ArtIA]

Madam Chair, members of the committee, thank you for the invitation.

I represent ArtIA, a group of cultural organizations, research centres and artists. Our goal is to understand the impact of artificial intelligence on the arts through action research. We have just released a report after more than two years of work, and an upcoming report will be tabled soon.

Silicon Valley's technology development model has been failing for decades. In the 1980s, it was already known that this model only served shareholders, not human beings. Today, artificial intelligence amplifies this problem. Artificial intelligence giants are funded to the tune of hundreds of billions of dollars. This difference in scale creates dominant positions that are hard to counterbalance.

Numerous studies and surveys, even some conducted by this committee, have shown negative effects of certain social platforms on the population, particularly on adolescents and marginalized people. Risks of political bias in referral systems have also been demonstrated. These same risks and biases are maintained in the different platforms of generative artificial intelligence, since their training bodies draw directly on the data produced in those networks.

We don't have a technology problem per se; we have a digital feudalism problem. Silicon Valley aspires to our dollars, data and talent.

Our conclusions are clear: This model optimizes shareholder profitability first, not the public interest. Without safeguards, ethics and cultural specificities become externalities, leading to cultural standardization. The social costs are borne by the communities, as well as by governments, which leads to cultural uniformity, especially among linguistic and cultural minorities. Without cultural sovereignty, there can be no economic sovereignty, since nothing would distinguish us as a society.

Silicon Valley controls all our production tools. Are we also letting them take control of artificial intelligence? That's the question we're asking.

Artificial intelligence is the next frontier of technological colonization. Here is what our findings reveal.

First, we see the exploitation of creators by dominant artificial intelligence models that are driven by data stolen from creators, which marginalizes francophone communities, aboriginal communities and other cultural minority communities, such as Acadians. This is unacceptable.

Second, there is the threat to cultural diversity. Because toxic algorithms homogenize cultures, our different languages are threatened.

Third, given the pace of development, accelerated exit cycles impose technical and cultural standards before public deliberation. Without adaptation laboratories, linguistic minorities, once again, don't have the time to adapt their tools, interfaces and data sets.

Fourth, we're experiencing a Napster 2.0 moment: Without timely intervention, we risk repeating the music industry’s mistake—ceding control of our cultural data and creative tools to foreign platforms.

Fifth, harnessing digital commons—in other words, what we've all created on the Internet—is at the heart of artificial intelligence models owned by digital giants.

The artists we work with don't seem to be afraid of technology. Instead, they're afraid that they won't be able to survive financially in a future dictated by technological giants.

ArtIA proposes government investment in laboratories for experimentation, training and production of artificial intelligence in the field of culture that remain owned by the Canadian cultural sector and are governed as a digital community.

These include laboratories that are experimental spaces where cultural communities design their own artificial intelligence tools, adapted to Canadian French, indigenous languages and other minority languages.

We also propose cultural data trusts, infrastructures that are sovereign and that protect and value our cultural data, whose governance rules and terms of access are decided by the communities and by the artists.

In addition, we favour training programs that enable artists to master artificial intelligence rather than be dependent on it, thereby preserving their creative autonomy and our creative autonomy.

In our view, this is an exportable model. If we succeed, we'll create a culturally responsible AI ecosystem model that can be exported to other social sectors and countries. As a result, Canada is positioning itself as a world leader.

I'd like to point out that Canada has already demonstrated its ability to implement innovative public policies. The 2017 Pan-Canadian Artificial Intelligence Strategy, the world's first national artificial intelligence strategy, has allowed Canada to become a global leader in artificial intelligence. This strategy has created hubs of excellence, such as Mila in Montreal, which has attracted international investment and positioned Canada as a key player in the ethical development of artificial intelligence.

ArtIA is part of that tradition. We want to develop laboratories for our culture by extending this visionary approach to the cultural sector, which will protect our diversity while creating economic opportunities.

Artificial intelligence is already transforming our creative industries. Acting now means choosing to protect our cultural diversity rather than going through homogenization.

The Canadian government invested $2.4 billion in artificial intelligence, but none of those funds went directly to culture, one of the most vulnerable economic sectors to—

[The Chair, Liberal, Lisa Hepfner]

Thank you, Mr. Ducharme.

Next, from OCAD University, we have Kelly Wilhelm.

You have the floor for five minutes.

[Kelly Wilhelm, Head, Cultural Policy Hub, OCAD University]

Thank you, Madam Chair.

Thank you, honourable members of the committee, for the invitation to appear today. My name is Kelly Wilhelm. I am head of the cultural policy hub at OCAD University. Prior to that, I spent about 20 years in the federal policy and funding system around arts and culture.

I want to start by breaking down a false dichotomy that we often hear in discussions around the creative industries and AI. That dichotomy is between two sides. On the one hand, there are those who argue that private sector development of AI must be allowed to continue unfettered and unregulated or the country's capacity to innovate will be at risk. On the other side, we often hear those who argue that government regulation and guidelines are essential to preserve the country's culture and identity. You can see the split clearly in the government's 2024 consultations around AI and copyright and also internationally in any policy discussion that addresses the use of copyrighted works in the development of AI systems.

Here's the reframing that I want to propose to this dichotomy. Government action that values and invests in creative and cultural professionals and their IP does contribute to innovation. It does not stifle it. There's no question that AI impacts creative IP, copyright and labour. It has already disrupted the value chains on which the creative industries, their companies and their workers depend. These are critical issues, and collective work on them by industry, researchers and policy-makers must continue.

You've heard already in these meetings that the creative industries are asking for transparency and fairness in how the content to which they own the rights is used in AI systems and in identifying AI-generated content for what it is. They're also asking for a seat at the table where AI policy decisions that affect them are being made. This is a point that I would echo and support. That's because, while AI is affecting the cultural industries, as your question and your study suggests, the effect is not one way. That goes back to my reframing. Like other technologies before it, creative industries are using AI to do what they do best—create, innovate and tell Canada's stories.

In the cultural policy hub's convening with creative industries over the past two years, we've heard that many are using AI and building AI tools that help them to run their businesses—the back end—and in their creative work. They use AI for many purposes, including creation, production, distribution and marketing. They use it to reduce technical, financial and environmental barriers for all sorts of creators, whether those are small and medium-sized enterprises, indigenous creators or even those working in rural and remote communities. They use AI to maximize the value of their IP, to find new markets and to build out fan bases in global markets. They use it to protect the sovereignty of the data they hold, which is cultural data, and to develop new protocols and ownership models. Artists, as we know, work with AI to create the entirely new and unexpected, very often in collaboration with tech companies and product designers in the private sector.

Creative industry leaders are asking how Canada can shape AI in a way that protects and enhances our culture, creativity and sovereignty. Four key themes keep coming up in our conversations with those leaders. I want to share them with you today.

The first is around sovereign and accessible data. Creative industries, like other industries, need access to data, including government datasets, to develop effective, Canadian-made AI tools and solutions. Governments should be investing in Canadian-owned and Canadian-governed public data infrastructure—an AI commons, if you will—alongside its investments in private companies and AI compute.

The second area that keeps coming up is skills training and education. Canada needs a national AI strategy for education from K to 12, post-secondary and in professional sectors like the creative industries. The strategy should prepare Canadians to harness new AI tools and technologies to remain competitive in the evolving job market. This is a concern the creative industries share with other industries.

The third area, which I'm sure you'll hear about again today, is regulation. Most in the cultural sector are calling for simple and clear transparency requirements, harmonized standards and shared principles in order to avoid uncertainty, fragmentation and confusion in this marketplace.

The last point is cultural sovereignty and creative ambition. This moment of economic, technological and geopolitical change is the right time for Canada to be ambitious in this space, not to draw back but to think big and invest in the full diversity of creators and creative SMEs that drive our creative economy in this country. Canada has long recognized the importance of our creative industries to our cultural sovereignty. It's why we have the tools and the successful industries that we do in film, TV, music and others.

AI is fundamentally a homogenizing tool. It is about recording and reproducing existing patterns found in the data that is fed into the machine. It does much of the cognitive work for us, replacing the human work of synthesizing and considering patterns that is essential to creating new ones.

Let's be clear: AI-generated content is not a replacement for original creative IP.

[The Chair, Liberal, Lisa Hepfner]

Thank you.

[Kelly Wilhelm]

Right now we need that creative IP as an engine to drive not just our economic sovereignty but our cultural sovereignty as well.

Thank you.

[The Chair, Liberal, Lisa Hepfner]

Thank you, Ms. Wilhelm.

I will now give the floor to Patrick Rogers from Music Canada.

You have five minutes.

[Patrick Rogers, Chief Executive Officer, Music Canada]

Thank you.

My name is Patrick Rogers. I am the CEO of Music Canada. Music Canada is the trade association of Canada's major labels—Sony Music Entertainment Canada, Warner Music Canada and Universal Music Canada. Our companies have offices full of Canadians making Canadian music for the Canadian market and the world. They are very excited about what AI could mean for the future of music.

For my members and the artists they invest in, AI as a tool can help human artists elevate their creativity, find efficiencies in the recording process and help our industry protect its IP. However, having faced an existential technological crisis in our lifetime, the music industry has some lessons learned that shape our advocacy to government.

As you know, not that long ago, home Internet created the possibility for peer-to-peer file sharing and Napster and other sites decimated the industry, but it wasn't the Internet that nearly killed the music industry. It was copyright piracy and the public's willingness to become pirates. It was the breakdown of our public understanding that we—I mean all of us—were stealing. I worry about that happening again today, or tomorrow, with AI, but let me first share some examples of AI at its best.

The team at Universal Music Group and Brenda Lee, keeping the original music and background vocals, replaced Brenda Lee's original lead vocals in Rockin' Around the Christmas Tree with a newly translated Spanish vocal, created using an ethically trained AI model derived from her voice and fully approved by her. After Randy Travis suffered a devastating stroke in 2013, leading to the loss of his voice, he and Warner Music Nashville used AI to help him record the song Where That Came From. Randy and his team worked on vocals and put human touches on every note of the song, which would not have been possible without AI. Universal Music Group's collaboration with ethical AI company Endel is enabling UMG artists and labels to create AI-powered soundscapes to enhance daily activities like sleep, relaxation and focus.

What's the problem? Each of these examples is done in collaboration with the artist involved. It's still their artistry. It's still their talent. Their rights are embedded in the project. That's not how the most popular song-generating services work now. They aren't licensed by artists and rights holders, and they aren't paying for the inputs they are trained on or the outputs that compete with those very artists. We are in the Napster era of AI in the marketplace. We need to get to the iTunes stage so that we can get to streaming. To do that, we need to do a couple of things.

One, uphold copyright law, recognizing that the training of AI models on the music of artists engages their copyright. Done without their permission, it is infringement—full stop. Upholding copyright also means that you should not cave to demands for sweeping “text and data mining” exceptions. Everything you need to know about this is in its name. It's not called the “hopelessly searching for human enlightenment” exception. They call it mining. What do we mine for? Valuable things. Where do we mine? Where we know there are valuable things. No data is more valuable than the catalogue of our favourite musicians. As much as some generative AI companies will tell you that they are teaching the machines everything, they don't want to train their algorithms on your brother's college band. They want Sgt. Pepper's and Pet Sounds and Folklore.

Two, we want to make AI companies keep records of their inputs. We want AI companies to track and disclose to rights holders what they train their models on. The proponents that say AI should be able to steal everything will claim that it isn't possible, but if we're going to unlock human consciousness with AI, shouldn't it be able to write a bibliography?

Three, we want the federal government to get serious about deepfakes, not just the worst kinds—the sexually explicit ones or the ones that interfere in our elections—but all of the harmful ones. They put all Canadians, including our kids, at risk. What started as Connor McDavid appearing to smack talk his opponents and teammates two years ago has turned into babies interviewing their dogs, which I know aren't real but I can't tell you why. These videos make us laugh, but the technology could easily be used to make us cry. What if someone used it to ruin your career? What if someone used it to ruin my five-year-old's budding social life? What's illegal on paper should be illegal online. Putting your words in my mouth is not free speech.

Fortunately, AI will also be part of the solution. It's why our industry is working with tech to combat this problem. Sony Music's work with Vermillio's TraceID tool helps protect artists in real time against digital replicas, but we need governments to make it clear that unauthorized and harmful deepfakes are illegal. There are models for Canada to look at, such as the U.S. NO FAKES Act, which has bipartisan support and is backed by both the music industry and platforms like YouTube.

These three recommendations will lead to Canada being at the forefront of AI innovation while still protecting the Canadian cultural sector.

Thank you.

[The Chair, Liberal, Lisa Hepfner]

Thank you, Mr. Rogers, and well done taking my cue.

Next, we go to Jennifer Brown from SOCAN.

It's over to you for five minutes.

[Jennifer Brown, Chief Executive Officer, Society of Composers, Authors and Music Publishers of Canada]

Thank you very much.

As CEO of SOCAN, I am pleased to appear before this committee to speak to the effects of artificial intelligence on the music industry.

In 2025, SOCAN is celebrating 100 years as Canada's copyright collective for songwriters, composers and music publishers. We are proud to represent the rights of more than 200,000 direct members from coast to coast to coast. SOCAN collects licence fees for the public performance and reproduction of music, matches those uses to rights holders and then distributes the royalties to our direct members and rights holders from across the world.

We are only starting to uncover the full potential and implications of AI in the creative sphere and to understand what role it will play in shaping the future of the cultural landscape in Canada. AI presents a turning point for the music industry. We believe that with proper safeguards and an appropriate copyright framework, this technology provides tremendous opportunities to support and enhance human creativity as a tool that allows Canadian creators to continue to tell their stories, reflect on who they are and contribute to Canada's identity and values.

However, without the appropriate balance in place, the current state of AI presents a challenge for our members. Canadian works are being stolen and scraped by AI companies to train their models, without any compensation to creators. These AI models can then output a complete song in response to a single prompt, with that AI-generated song potentially replacing the work of Canadian creators.

A global study conducted by CISAC, which is one of the world's largest networks in the creator sector, estimates that under the current market conditions of wholesale theft, up to 24% of music creators' revenues are at risk of disappearing. This presents a real threat to the sustainability of the Canadian music industry.

Meanwhile, the companies behind these AI models have publicly stated the importance of using high-quality, human-created music to develop their products, going so far as to say it is essential. What is not essential, apparently, is making sure that creators are paid for their important contributions. Instead, human creators are currently fuelling advances in AI models without sharing in the benefits.

We believe a successful AI approach will value and compensate human authorship, respect the policy objectives of the Copyright Act and lead to a vibrant licensing market where the benefits of AI are shared with those who are vital to its development.

Looking back at the online streaming we've gone through in the last 10 years or more, similar arguments were made about compensation and licensing being impossible, but a mature licensing regime has formed and creators and streaming services both have benefited in the past decade. Respect for copyright does not stifle innovation. If you stream music on your smart phone, you have proof in your pocket that compensation for creators and technical innovation can successfully coexist. The adoption of AI can also be done in a way that respects creators and incentivizes human expression.

We have two positions that we would like to put forward to the committee.

First, we strongly oppose new copyright exception. AI companies should not be permitted to exploit creators' works without obtaining consent and providing credit and compensation. A TDM exception would not facilitate growth in either the creative sector or the technology sector. While there is no evidence to suggest that a TDM exception is necessary to maximize investment in the AI sphere, it would certainly deprive creators of the economic benefits of their works.

Second, we urge you to ensure AI companies are transparent about the works they use to develop their models and that AI-generated outputs are clearly labelled. AI developers must be required to disclose which copyright-protected works are ingested and stored in their datasets. Without such transparency, rights holders are unable to negotiate on a level playing field and cannot prove when their works are used. Further, mandatory labelling of AI outputs would mean that the public can then make informed choices about the type of content they consume.

Thank you very much for your time.

[The Chair, Liberal, Lisa Hepfner]

Thank you.

Finally, from Music Publishers Canada, we have Margaret McGuffin.

You have the floor for five minutes.

[Margaret McGuffin, Chief Executive Officer, Music Publishers Canada]

Thank you.

Good afternoon, Madam Chair and members of the committee.

My name is Margaret McGuffin and I am CEO of Music Publishers Canada. I am here to advocate for the ethical and transparent development of AI.

Music publishers discover and develop Canadian songwriters and have made significant investments in the vast majority of songs and scores that are heard every day on radio, on streaming services, in video games, in film and television productions and on new emergent platforms around the world.

In the music space, AI has the potential to support the valuable work of human creators, which in turn enriches Canadian culture and society. Our members are already leveraging the benefits of this new technology. Songwriters are using it in the studio, and our members are using it to scale their operations.

Unfortunately, the music industry has also seen mass theft of copyrighted-protected songs by AI companies, both on the input side for the purpose of training AI models, and on the output side, with the development and publication of unlicensed generative AI models. This poses serious risks for Canadian creators and the companies that invest in them.

Strong copyright ensures that MPC's members, songwriters and composers maintain control over their music and receive the fair compensation they deserve.

When an AI company uses music that has been scraped or captured from the Internet without authorization, it prevents rights holders from controlling and realizing value for the use of their works. The development and commercialization of unlicensed AI model inputs and generative AI outputs are already creating serious market distortions and raising concerns about fair competition.

MPC works with the International Confederation of Music Publishers. A recent Billboard magazine story highlights evidence collected by ICMP over the past three years showing that many of the world's biggest tech companies have scraped copyright-protected music created by millions of songwriters, composers and artists to train generative AI systems, without permission or licensing.

To put this in perspective, nearly every song ever written by a Canadian songwriter has already been scraped and is already stolen by these AI companies without consent, credit or compensation. Imagine that someone accessed your paycheques without your permission and that behaviour was normalized. That's what's happening to songwriters.

This extensive non-compliance with copyright laws in turn leads to serious negative economic impacts. Copyrighted works—our songs—add value to AI models. To derive fair value for the use of this copyrighted material, the music publishing industry, which includes SOCAN and CMRRA, routinely grants licences to technology companies. AI developers should be no different. The emerging market for licensing music to AI developers should be encouraged, including requiring them to disclose and maintain records of all their data.

In conclusion, MPC believes the Canadian government must reject any calls for watering down the copyright system with a text and data mining exception. We've already heard about that today. Music rights holders must be able to control and realize value for the use of their songs. It is imperative that Canadian regulators and the U.S. government approach generative AI in a manner that respects creators and incentivizes human expression. This will be beneficial for creators and for Canadians as a whole.

I look forward to answering any of your questions.

[The Chair, Liberal, Lisa Hepfner]

Thank you very much, Ms. McGuffin.

We'll now turn to questions from members, starting with Mrs. Thomas.

For six minutes, you have the floor.

[Rachael Thomas, Conservative, Lethbridge, AB]

Thank you very much.

Mr. Ducharme, I'm just curious. Does your organization work with digital-first creators at all?

[Marc-Olivier Ducharme]

It's with all kinds of creators, so it's all the forms of art, from dance to visual arts. We also work with musicians, with sound artists and with theatre companies. It's really with all ways of life and types of art.

[Rachael Thomas, Conservative, Lethbridge, AB]

Okay. Cool.

If digital-first creators are working with you, are there helpful ways in which they're using AI in order to overcome the traditional barriers of broadcasting?

[Marc-Olivier Ducharme]

Yes. Most of the artists who work with us actually use AI on a daily basis, whether it's integrated in their tools, like Adobe Photoshop, which has already integrated AI into their products, or the camera on their phone, which already integrates AI into the process of taking a picture. They use AI. Whether they want it or not, they do.

We all use AI right now. It's not that.... It's very helpful. Also, a lot of artists want to do research and development around AI and how it can be used to better their practice.

The problem is really not in the use of AI by artists. The problem is really in the value that is transferred to American companies through the process of value capture.

[Rachael Thomas, Conservative, Lethbridge, AB]

Sure. I think one of the things you talked about in your opening remarks, and that I read also on your website, was the importance of protecting digital sovereignty. To that end, Bill C-18 is the Online News Act. Are you familiar with it at all?

[Marc-Olivier Ducharme]

Yes.

[Rachael Thomas, Conservative, Lethbridge, AB]

Okay.

Bill C-18 was put in place and, through its process, Bill C-18 resulted in Google coming forward and creating a hundred-million dollar fund. There was then a cohort put in charge of that money. News businesses would apply to the fund and then be awarded dollars.

As a result, when they are awarded those dollars, they are also signing over their rights to the information they're producing, which means that Google can then go in, scrape that information and use it however it wishes. Because these news businesses have signed off on Bill C-18 and the Google dollars, they're subject to Google and whatever Google wants to do with their information.

This is the direct result of over-regulation by this government, so my concern is this. We have digital-first creators in this country who are generating fantastic content. We have artists in this country who are doing great work. When the government overreaches, it stagnates and hurts the industry. Where is that line between putting legislation in place that is going to facilitate further success versus hindering it?

[Marc-Olivier Ducharme]

It's an interesting political point you make, but my fellow members here will be able to answer you on the policy side.

What I think is needed is investment in a public AI, in public infrastructure and an infrastructure that is common, which is led, managed and owned by artists and arts organizations in Canada. It's not necessarily a question of regulating. You're asking me how the government can regulate, which is not my expertise. What I can tell you is that the model needs to be managed in a different way right now.

[Rachael Thomas, Conservative, Lethbridge, AB]

If it's managed in a different way, but without government regulation, what does that look like?

[Marc-Olivier Ducharme]

I am not a policy expert. I would not be able to inform you on how the government can really create policy to protect the artists. If we just choose to give the power to American companies, we're not going to win. Basically, it's that.

[Rachael Thomas, Conservative, Lethbridge, AB]

Absolutely, but again, with Bill C-18, that's exactly what happened with this government legislation. It gave more power to Alphabet, to an American company, Google, to scrape data and use it however it wishes.

[Marc-Olivier Ducharme]

They are stealing it anyway.