A Look into Indonesia’s AI Roadmap

What the National AI Roadmap Means for Intellectual Property

Last year, the Ministry of Communication and Digital Affairs of Indonesia (Komdigi) published Indonesia’s National AI Roadmap and a draft of the Guidelines on AI Ethics. Both are meant to anchor a forthcoming Presidential Regulation. Together they set the country’s direction on artificial intelligence. Yet neither says anything specific about what artificial intelligence means for intellectual property law in Indonesia.

That silence matters. It leaves a basic question open. How do copyright, patent, design, and trademark rights apply when a machine makes the work, the invention, or the evidence? For rights holders, this is practical, not academic: who owns AI output, how is it protected, how is it enforced?

AI Policy With Opportunities to Strengthen Intellectual Property Provisions

Indonesia’s intellectual property system rests on key statutes like the Copyright Law, the Patent Law, the Industrial Design Law, and the Trademark Law. None was drafted with generative artificial intelligence in mind. The National AI Roadmap does not bridge that distance. The same gap reaches adjacent laws the reader will recognise; the Personal Data Protection Law, the Consumer Protection Law, the competition law, and the Electronic Information and Transactions Law.

Without explicit AI provisions, uncertainty falls on three fronts: ownership, protection, and enforcement of AI-generated rights. In its submission on the National AI Roadmap, SKC Law set out where the statutes leave artificial intelligence unaddressed. A consistent fix can be built across the IP framework, and the submission shows how.

Where AI Meets Intellectual Property Law

Copyright: Who Is the Author of an AI Work?

The Copyright Law (Law No. 28 of 2014) does not expressly govern works generated by artificial intelligence. It leaves open whether AI output is a “work” at all. If it is, who counts as the “author” – the human user, the developer of the AI, or no one? The Dewan Perwakilan Rakyat (DPR) / House of Representatives is now reviewing a revision of copyright law in Indonesia, prompted by the public dispute over royalties for songs played in public spaces. Indonesia’s Directorate of Copyright and Industrial Design has signalled it will look to provisions used abroad. One example is the Tennessee ELVIS Act, which protects the voice, image, and likeness of public figures from AI imitation. A workable rule would separate AI-assisted works from fully AI-autonomous ones, rewarding human creativity without over-extending. Training data is the other open question, and it should run on transparency and fair licensing.

Patents: Can an Inventor Be a Machine?

The Patent Law (Law No. 13 of 2016, as amended and most recently consolidated under Law No. 65 of 2024) sets the familiar tests of novelty, inventive step, and industrial applicability. How those tests apply to an AI-generated invention is unsettled. So is the threshold question: can an AI be named as inventor, or is inventorship limited to people and legal entities? Patent law in Indonesia will need to take a position. Artificial intelligence also has a constructive role on the examination side. Patent examiners at the Directorate of Patents of Indonesia, part of the Directorate General of Intellectual Property (DGIP), can use AI for prior-art searches and substantive examination. Clear usage guidance would capture that efficiency while preserving quality and accountability.

Industrial Design: Novelty and Mass-Filing Risk

The Industrial Design Law (Law No. 31 of 2000) raises the same ownership question; whether an AI-generated design can be registered, and in whose name. Because design protection turns on novelty, the regime is exposed to a specific abuse: AI-driven mass filing that can crowd out legitimate human designers. However, the same technology cuts both ways. Artificial intelligence can detect design similarity, support consistent examination at the DGIP, and help the Indonesian National Police (Polri) and civil-servant investigators (PPNS) assess infringement.

Trademarks: AI Evidence in Enforcement

Trademark enforcement increasingly relies on artificial intelligence, such as image recognition to flag counterfeits and automated monitoring of online marketplaces. Trademark law in Indonesia (Law No. 20 of 2016) does not yet address how evidence generated this way is treated in opposition, cancellation, and infringement proceedings. For rights holders, the practical question is admissibility: will an AI-detected match stand up, and against what reliability standard?

Aligning With International Standards

Indonesia is not starting from scratch. The World Intellectual Property Organization (WIPO) has an active workstream on artificial intelligence and intellectual property, providing a reference point for any domestic reform. Aligning with these international developments can strengthen cross-border IP protection and enhance Indonesia’s appeal for innovation and investment.

What This Means for Businesses

For a business creating or policing intellectual property with artificial intelligence in Indonesia, the exposure is concrete. In practice, ownership of AI-assisted output cannot be assumed, because the statutes do not yet say who holds it. Enforcement built on AI-generated evidence carries an admissibility risk. That risk will not resolve until the Courts and the DGIP set a standard. Importantly, the window to act is open now; through the Copyright revision now before the DPR and the consultation that will shape the Presidential Regulation.

There are two practical steps to start taking now. Rights holders should document the human contribution to AI-assisted work, and treat AI evidence as a reliability question rather than a given. Both leave them in a stronger position whichever way the rules land.

Strategic Recommendations

1. Add AI provisions to the Copyright revision now in progress. That revision is the nearest legislative vehicle. Provisions added here should stay consistent with the other three IP statutes, not stand alone.
2. Build a cross-cutting AI adjustment across the IP statutes. A single set of AI provisions spanning patent, industrial design, copyright and trademark law would prevent disparity and inconsistent interpretation. It would settle the recurring questions: ownership, authorship and inventorship, and registration of AI-generated works.
3. Set reliability standards for AI-generated evidence. Courts and administrative bodies can develop guidance so that AI-detected infringement evidence is admissible against a defined standard. Today it is decided case by case.
4. Use AI in the examination and enforcement side, with guardrails. Prior-art search, design-similarity detection, and marketplace monitoring all benefit from artificial intelligence. Clear usage guidance lets DGIP and enforcement bodies capture that efficiency while protecting quality and accountability.
5. Engage the multi-stakeholder consultation. Government, academia, industry, and rights holders each hold part of the picture. Businesses with a stake should treat the consultation as the moment to shape it, before the Presidential Regulation fixes the position.

Connect With Us

SKC Law advises clients across the full intellectual property lifecycle in Indonesia; filing strategy, portfolio management, prosecution, advisory, and enforcement. We advise rights holders on the questions artificial intelligence now sharpens, from ownership of machine-assisted work to the evidentiary side of online enforcement, where Indonesian courts have already begun to accept electronic and social-media evidence in trademark matters.

For enquiries on artificial intelligence and intellectual property law in Indonesia, contact our team at enquiries@skclaw.id.

Follow us on LinkedIn for updates on intellectual property in Indonesia.

This content is provided for general information only and does not constitute legal advice. For advice on specific matters, contact enquiries@skclaw.id.