By Brontë H. Lacsamana, Reporter

CREATORS from artistic, literary, musical, and audiovisual industries are pushing for a better regulatory framework behind the use of artificial intelligence (AI) models. As their usage increased over the past five years, there have been increasing concerns over the loss of revenues due to unauthorized use of creative works by these AI models as well as AI-generated art competing against the human-made.

Emerson G. Cuyo, director of the Intellectual Property Office of the Philippines’ (IPOPHL) Bureau of Copyright and Related Rights has reported that “guidelines on the copyright registration of literary and artistic works that contain material generated by AI” are now being crafted.

As of now, creative works must be the work of a natural person to be eligible for copyright protection.

“We rely heavily on the disclosure of the applicants,” said Mr. Cuyo in a talk last October on copyright. “What’s for sure is that works that are wholly generated by AI do not pass the requirement for copyright protection under our law. But for partially generated works, that’s where the confusion is, and that’s where guidance will come in.”

MARKET FOR AI EXPLODING
While no creative body in the Philippines has data on the size of the market for AI-generated works today, the International Confederation of Societies of Authors and Composers (CISAC) commissioned PMP Strategy to do a study on global companies’ increase of interest in Generative AI.

Generative AI (Gen AI) refers to generative models that produce text, images, and videos trained on underlying patterns and structures of data fed into the system. Language prompts then allow the models to produce new data based on the input. It was made popular by ChatGPT.

The study revealed that the market for music and audiovisual content generated by AI will increase exponentially in the next five years, growing from around €3 billion now to €64 billion in 2028.

Meanwhile, the revenues of Gen AI services in music and audiovisual industries are projected to reach €9 billion in 2028, up from €0.3 billion now.

Gadi Oron, CISAC’s director general, said in a statement for their global collections report in December that “new challenges brought by artificial intelligence technologies threaten the future of the creative sector.”

On the ramifications of widespread Gen AI use, he explained: “It is not simply another means of distributing creative works; it is a technology that appropriates and replicates them. Without proper safeguards or data transparency it risks undermining the very foundation of creative value.” 

CISAC represents 228 collective management organizations, 116 countries, and over 5 million creators across music, audiovisual, drama, literature, and visual arts, according to Mr. Oron.

Of these industries, royalty collections for music totaled €12.59 billion in 2023, with audiovisual content in second place at €730 million. This makes these two subsets of creators most susceptible to revenue loss in light of the potential dominance of Gen AI services. 

Unlicensed Gen AI could divert up to 25% of creators’ royalties, equivalent to €8.5 billion annually, if left unregulated.

For ABBA singer-songwriter and composer and CISAC president Björn Ulvaeus, authors’ societies delivering record royalties to creators worldwide may be a cause for celebration, but the rise of AI warrants serious attention.

“The advent of artificial intelligence signals a profound shift for our sector — proof that progress and disruption can exist side by side, and that the future of creativity will be shaped by how we reconcile both realities,” he said in a statement.

Locally, based on stakeholder feedback, IPOPHL revealed that the music, audiovisual, and publishing industries are the sectors most concerned by AI use.

VIGILANT CREATIVES
In 2023, The Beatles released “Now and Then,” billed as their last song. The track featured the voice of the late John Lennon, which an AI program extrapolated from an old recording. This is an example of AI use as a tool that draws from a small, curated subset of information.

On a wider scale, it is being used to generate entire videos and songs, using data extracted from creative works found all over the internet — without permission from creators.

The most recent example is an unfinished local theater production titled Moudifa the Musical, which presented a proof of concept for AI-generated music and lyrics back in November.

BusinessWorld attended the preview for 14 songs, where entrepreneur and budding writer-composer Margarita Marquis proudly claimed that they were written entirely by herself “with the aid of AI.”

“At the beginning, a hundred songs I created with AI. I said, ‘Okay, I will use AI, no soul,’ kasi sabi nilang lahat walang soul daw, diba (because everyone says that it has no soul, right)? So I created [it] like this. And then, at 3 o’clock in the morning, pwede na (it was okay),” she said, when we asked what her creative process was.

“After 50 songs, AI was following my feelings, my soul! I said, ‘Okay, AI, can you cry?’ AI can cry!” Ms. Marquis explained at the press conference. “Did you hear the lyrics? It is not AI. It is from my soul.”

Luna Griño-Inocian, a playwright and librettist, noted in reaction to this that AI ultimately “gets emotions and feelings from somebody else,” using original work by other artists as a template.

“When your ideas are posted online, they get eaten up and used,” she said in a Zoom interview in November. “You can spot if something looks or sounds like something else because AI eats up publicly available works by human beings and throws that back at you.”

Composer and scorer Vincent De Jesus said in a video call that claims of an AI future where performers will be obsolete will surely be protected by a musicians’ equity, such as when unions on Broadway protested over Here Lies Love using only a DJ and canned music, and no live musicians.

He explained that the Filipino creatives can be “just as vigilant” if a company were to open and mount shows fueled by AI.

“The industry and the audiences will not stay silent. Coming from the pandemic, we’ve only just begun reaping the fruits of our labor,” Mr. De Jesus said. “Starting 2023, left and right we’ve had so many original plays, musicals, student productions. At the end of the day, the audience will be the judge.”

DEMISTIFYING AI
Early in 2025, a band named The Velvet Sundown, with about 500,000 monthly subscribers gathered over the course of a month, was revealed to be an AI band.

“The Velvet Sundown is a synthetic music project guided by human creative direction, and composed, voiced, and visualized with the support of artificial intelligence,” said its spokesperson Andrew Frelon in a post online. The generative AI program Suno was used to create the virtual act, complete with consistent songwriting.

Trina Belamide, a songwriter and board member of the Filipino Society of Composers, Authors, and Publishers (FILSCAP), explained at a Linggo ng Musikang Pilipino (LMP) event in July that creators and founders of programs like Suno and Udio “make art a commodity.”

“In their minds, they help people so that making art will no longer be difficult. They think, who wants to suffer? But they miss the whole point, because that’s exactly what we love to do as artists,” she said. “If you’re an artist, you enjoy the process of creating.”

In an e-mail interview with BusinessWorld, cultural writer and musician Mariah Reodica explained that, although there is only so much she can do with AI in her personal practices of writing, music, and video, the way forward in general is to demystify the technology.

“I see these models as more of a tool than anything. They’re efficient for tasks that can be automated like parsing data or proofreading, but efficiency doesn’t necessarily mean that the quality of output is good. I always have to step back in at some point in the process to validate or reject what these produce. I have the final say,” she said.

Citing Meredith Walker, president of global communication non-profit Signal Foundation, Ms. Reodica added that creators need to “take a step back from taking the term ‘artificial intelligence’ at face value.”

“We can begin to assuage worries and technophobia by beginning to understand it. Artists in conversation with digital rights advocates in conversation with policymakers in conversation with open-source programmers in conversation with educators in conversation with everyone else,” she said.

Award-winning screenwriter and National Artist for Film Ricky Lee said at a book launch in November that human-made stories foster empathy, which AI can never replicate.

“Sa pamamaraan ng storytelling, natututo tayong makipamuhay sa mga karakter at maintindihan ang bawat pagsubok natin bilang tao. (Through the use of storytelling, we learn how to live with the characters and understand every struggle we face as humans). AI would not get that,” he said.

“I don’t think it can ever supplant our own voice, our conscious storytelling, the in-betweens of words,” Mr. Lee added. “It’s a tool, but we have to have agency over it.”

On social media platforms like Facebook, many accounts spread AI-generated videos that foster disinformation and attempt to mimic real human situations and emotions to garner clicks, likes, and shares. Telltale signs include the watermark of Gemini, Google’s generative AI program, in the corners of the frame; misspelled or distorted text; and unusual or inconsistent movements and facial expressions of people in the video. Though according to some experts, these telltale signs are fast disappearing as the AI models improve.

OPEN SOURCE
For Christina Lopez, an Ateneo Art Award-winning visual artist who uses an AI model to produce images in her works, the tool can be used in the same way DJs sample tracks or chefs make use of a bottled sauce.

“It’s basically a component. If I didn’t already have an image in mind or a clear direction to take it, AI wouldn’t be useful,” Ms. Lopez explained. “The work I show is not the AI-produced thing in itself.”

Her contemporary art practice, spanning sculpture and video, directly deals with the power relations and implications behind image-making.

Both Ms. Reodica and Ms. Lopez told BusinessWorld that open-source software has a friendlier model which champions creators and digital rights advocates. It refers to source code made available to the public, allowing anyone to view, modify, and distribute the software.

This, compared to the likes of OpenAI, which used to be an open-source company and is now for profit, and X (formerly Twitter), which owns all the user-generated data on the site.

“It’s alarming that AI tools’ development is being dictated by big tech conglomerates who aren’t transparent about their intentions or goals,” said Ms. Reodica. “We’re thinking that we’re gaining something from using a service at little to no cost like ChatGPT or Midjourney in exchange for our data. We don’t know what it’s being used for exactly.”

For Ms. Lopez, who began using AI tools in 2019 (training her own models and not relying on large-scale, internet-wide data), the shift from artist-centered to profit-centered work was evident.

One of the open-source tools she used was Style GAN, short for Style Generative Adversarial Network, which generates diverse, high-quality images. When it was acquired by the company Nvidia, the data was no longer usable.

“It’s kind of this weird negotiation where artists would worry about these companies stealing from them. Intellectual property rights now don’t even consider that, the fact that AI is not being built in the first place for artists,” she said.

This is why pattern recognition-savvy artists and creators these days are “more optimistic about smaller, open, user-based software built for the community.”

Similarly, Ms. Reodica said that creative works don’t have to be judged for whether they are AI-made or not. “I’m interested in work that values freedom, agency, and accessibility,” she said.

COPYRIGHT POLICIES
Musician and composer Myke Salomon told BusinessWorld in a video call that “no aspiring creative should be immediately crucified just for AI use,” and that anti-AI efforts should instead be redirected to regulation.

“When I started out, I used the computer to help me read and write notes and compose. I can relate on that aspect of it,” he explained. “I understand people’s need for a tool. At the end of the day, it boils down to using it properly and ethically to communicate your story.”

The CISAC and PMP study, among others that will likely be developed as AI continues to rise, aims to serve as a guideline for policymakers and legislative bodies.

Moving forward, at least on the Philippine level, the current IP Code is clear that “only a natural person (i.e., a human being) can be considered as an author of a copyrighted work,” according to IPOPHL’s Mr. Cuyo.

This means that an AI program on its own cannot be considered as the author and copyright owner of any work it generates. “It’s not a natural person,” he explained.

The IP Code as it is has no room for complexities, such as whether or not the human who is using the AI program is wholly or partly considered an author and copyright owner of the AI-generated work.

According to IPOPHL, the challenge in formulating policies is being both “responsive to technological trends and rooted in the existing legal framework.” Alongside this challenge is a positive — the office reported that copyright registrations in the Philippines reached a record high in 2024, with 6,552 IP certificates issued.

Last month, IPOPHL’s acting director general Nathaniel S. Arevalo said in a stakeholders forum that AI readiness is now “a priority issue all the way to the top,” with an AI strategy currently in development. The Bureau of Copyright and Related Rights is the department carefully studying the legal and practical ramifications of AI on copyright protection.

For visual artist Ms. Lopez, the hope is for intellectual property offices not to “chase after the technology that’s already being established.” The same goes for labor laws surrounding the AI creative tech sphere, with those providing the “creative fuel” of the Gen-AI content market not getting compensated, she said.

Aside from being an unregulated system, many creatives also see AI as “a looming threat that devalues human creativity,” as per the economic projections of the CISAC and PMP study.

Musician and cultural writer Ms. Reodica concludes that all the technological innovations forwarded by big companies are, in the end, “market-driven developments,” though neither “technophobia nor tech purism” will equip creators with what is needed to keep up.

“We need to understand AI. That way, we can work towards technological innovations that actually serve us.”