When AI processes Portraits: A Copyright and Ownership Perspective

Street photography and the use of human portraits have taken on new legal significance in the age of Artificial Intelligence (“AI”). Photographs captured without consent are no longer static images, but can be transformed, replicated, or monetized through AI systems, raising complex questions about copyright, moral rights, portrait rights, and ownership. Amid regulatory gaps under the current Law Number 28 of 2014 on Copyright (“Copyright Law”) and evolving approaches in the Copyright Bill or Rancangan Undang-Undang Hak Cipta (“RUU Hak Cipta”), consent and human creative contribution remain central to risk management. This article examines these issues from an Intellectual Property (“IP”) perspective and outlines their implications for individuals, businesses, and the creative industry.

Street photography has once again drawn public attention, particularly when portraits of individuals are captured in public spaces without consent. In practice, those photographs are frequently published on social media, offered as artistic works, or monetized for commercial purposes, often without the subject’s knowledge.

As technology advances, however, these issues no longer stop at ethical boundaries or consent in photo-taking. The emergence of AI has fundamentally altered the landscape. photographs taken without consent are no longer static works, but can now be processed, modified, or reproduced through AI into entirely new images detached from their original context. At this point, street photography intersects directly with issues of copyright, i.e., moral rights, portrait rights, and ownership of works in the AI era.

This article examines these issues from an IP perspective, using street photography as a starting point and expanding the discussion to portraits of individuals in AI systems, existing regulatory gaps, and the implications for individuals, businesses, and the creative industry.

Copyright and Portrait Rights in AI-Generated Photo Processing

Under Indonesian law, the Copyright Law provides legal protection for photographic works as protected subjects. Article 40 paragraph (1) expressly recognizes photography as a protected work, granting both moral and economic rights to photographers, as the creators, over their works.

However, when a photographic work contains an identifiable portrait, legal protection is subject to additional statutory restrictions. Article 12 of the Copyright Law stipulates that the commercial use of a person’s portrait requires the consent of the individual depicted, notwithstanding that the copyright in the photograph remains with the photographer. In other words, although the photographer holds copyright in the photograph, the subject retains legal rights over their face and personal identity, also known as portrait rights.

This issue becomes significantly more complex when photographs, particularly those taken without consent, are then used as input in AI systems. In practice, a person’s portrait can be processed by AI to generate a new image through facial modification, contextual alteration, or the creation of synthetic visuals that resemble the real person. Where the original photograph is also taken without consent, and then used as training data or as a prompt for AI, the legal question is no longer straightforward: does this still qualify as a permissible use of a photographic work, or does it involve a broader set of infringements, encompassing copyright, moral rights, portrait rights, and ownership rights?

Ownership of AI-Generated Works: Who Holds the Rights?

Another critical issue in this discussion concerns ownership of works generated by AI. As explained by Achmad Iqbal Taufiq, S.H., M.H., from the Directorate of Copyright and Industrial Design of the Directorate General of Intellectual Property (“DGIP”), practices across several jurisdictions, such as the United States, the European Union, and China, recognize copyright protection for AI-generated works, but determining who holds copyright in AI-generated works remains unclear.

In this approach, copyright is understood to subsist in the work itself, while ownership depends on the degree of human creative contribution involved in the use of AI, e.g., formulating substantive or complex prompting, making curatorial choices, or directing the creative outcome. The resulting work may be attributed to the human prompter, with AI functioning merely as a tool. Conversely, if the human contribution is minimal, copyright ownership may not vest in the prompter.1

This approach, however, becomes more complex when the AI input consists of a person’s portrait obtained without consent. In such circumstances, any ownership claim made by the prompter or AI user cannot stand independently, as there is a potential violation of the subject’s rights from the outset. Accordingly, even when an AI-generated work meets the threshold of human creativity, unauthorized use of a portrait may undermine the work’s legal standing from the outset.

This raises a recurring question among creative industry practitioners and businesses: does ownership rest with the AI user, the AI platform developer, or the individual whose face underpins the visual output? The Copyright Law has yet to provide a definitive answer, and the DGIP has acknowledged a regulatory gap regarding the legal status of AI-generated works.

How the RUU Hak Cipta Addresses AI-Generated Works and Portrait Rights

Unlike the currently effective Copyright Law, the RUU Hak Cipta, 22 November 2025 version2, expressly acknowledges the existence of AI-generated works and the legal risks arising from their use, including risks directly related to the use of portraits, personal identity, and individual likeness.

Article 63 of the RUU Hak Cipta recognizes individuals or legal entities as creators, insofar as they exercise creative control over the use of AI. Within this framework, AI is not regarded as a legal subject but as a technological instrument that supports human creativity. Applied to the context of street photography and AI-generated photo processing, this means that the individual who designs and controls the AI process may qualify as the creator and potential copyright holder.

Notably, the RUU Hak Cipta stipulates that any person who produces AI-generated works is subject to specific legal restrictions, both in terms of copyright protection and the protection of the subject’s personal interests. The restrictions include prohibitions against:3

Removing or altering information indicating the use of AI in a work;
Using AI to imitate the distinctive style of a particular author without authorization;
Generating works through AI that infringe moral rights;
Using or misusing an individual’s name, photograph, voice, or likeness, whether for commercial or non-commercial purposes; and
Creating misleading representations concerning the identity of an author through AI.

By situating portrait protection within the context of AI, the RUU Hak Cipta effectively extends the protection previously recognized under Article 12 of the Copyright Law. Under this RUU framework, using a person’s portrait as input to an AI system may constitute an infringement if done without consent and in a manner that infringes the subject’s personal interests, regardless of whether it will be used in commercial or non-commercial purposes.

From a risk perspective, this framework encourages a layered reading of the legal risks in AI-related use of portraits: an initial issue can arise when a photograph is taken or used without consent, and again when that same image is later fed into AI-driven processes. In this sense, AI does not constitute the source of the infringement itself, but rather forms a technological context in which the absence of consent or non-compliance with applicable legal restrictions may give rise to legal consequences.

Beyond substantive prohibitions, the RUU Hak Cipta also introduces administrative sanctions, ranging from written warnings and administrative fines to temporary suspension and termination of access.4 At this stage, the RUU Hak Cipta does not yet provide for specific criminal sanctions in relation to AI-generated infringements. However, it is important to note that the RUU Hak Cipta remains a legislative draft and has not been finalized. As such, its substance remains subject to further deliberation and potential revision prior to enactment, including the possibility that criminal sanctions may be introduced in future iterations of the bill.

Should the RUU Hak Cipta ultimately be enacted without specific criminal sanctions governing AI-related use of portraits, criminal liability will continue to be assessed under the existing framework of the Copyright Law, particularly Article 12. Until legislative clarity is achieved, the regulatory landscape surrounding the use of portraits in AI systems therefore remains transitional, underscoring the importance of careful adherence to existing consent and copyright requirements.

Conclusion

The use of human portraits in the AI era has transformed street photography from an ethical question into a legal issue involving copyright, moral rights, and ownership of works. In the absence of explicit regulation on AI-generated works, careful handling of portraits, proper consent, and a clear understanding of the boundaries of human creativity are crucial factors in mitigating legal risk.

For businesses and creative industry players, the key lies in ensuring a clear legal basis at every stage, whether through valid copyright ownership, proper consent from the individuals concerned, or adherence to the threshold for human creative contribution. Regulatory gaps do not imply freedom from liability; they call for more deliberate and responsible risk management in the use of AI.

For those seeking to navigate these issues responsibly, the Intellectual Property Practice Group at ADCO Law stands ready to provide strategic and practical legal guidance tailored to your business needs.

***

About ADCO Law:

ADCO Law is a law firm that offers clients a wide range of integrated legal services, including commercial transactions and corporate disputes in a variety of industry sectors. Over the course of more than a decade, we have grown to understand our clients’ industries and businesses as well as the regulatory aspects. In dealing with business dynamics, we provide comprehensive, solid legal advice and solutions to minimize legal and business risks.

From Upstream to Downstream, We Understand Your Industry. In complex transactions and certain cases, we actively engage with financial, tax, and environmental specialists, accountants, and law firms from various jurisdictions to add value to our clients. Our strong relationships with Government agencies, regulators, associations, and industry stakeholders ensure that our firm has a holistic view of legal matters.

ADCO Law is a Proud Member of the Alliott Global Alliance (AGA) in Indonesia. Founded in 1979, AGA is one of the largest and fastest-growing global multidisciplinary alliances, with 215 member firms in 95 countries. As a law firm, we also believe in regeneration. To stay abreast of business changes and stay relevant, our formation of lawyers is comprised of the top graduates from Indonesian and international law schools.