🔍 Read the full analysis: Why Indonesia’s AI Copyright Bill Has Meta, TikTok And Google Concerned on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Google, Meta and TikTok raised objections at an Indonesian parliamentary hearing on a proposed copyright-law revision covering AI and online news. The draft could require payments for some uses of news content and set rules for AI training, but its text is still being revised and no vote date has been announced.
Meta and TikTok joined Google in challenging parts of Indonesia’s proposed copyright-law revision at a parliamentary hearing in Jakarta on November 11, raising concerns about payments for news links and rules for artificial-intelligence training. The draft could reshape how digital platforms use publishers’ material and copyrighted works, but it is not law and lawmakers have not announced a vote date.
The bill would revise Indonesia’s Copyright Law, Law No. 28 of 2014. Parliament approved the revision as a legislative initiative on March 12, 2026, and its Legislation Body is now deliberating it with the government. The November 11 hearing included Google, Meta, TikTok, the Press Council and Satu Lisensi Indonesia, according to Kompas.
The companies’ objections cover different parts of the proposal. Google opposes payments connected to links and news snippets, a single collective licensing body, and mandatory licenses for AI training data. It also wants intermediary protections, known as safe harbor, retained. Meta argued that sharing a link is not the same as republishing an article, and said payment duties could reduce publishers’ traffic and income. TikTok said AI disclosure rules should not apply when AI is used only as an assistant, and also called for safe-harbor protection.
Google’s representative, Agung Pamungkas, said a levy on news links would leave the company with “no choice” but to stop crawling, indexing and distributing Indonesian news, Reuters reported. The company says licensing every item of data used to train AI models would be technically unworkable at web scale. It has proposed a text-and-data-mining exception with an opt-out mechanism such as robots.txt. These are the company’s arguments; the bill’s final terms remain unsettled.
Indonesia’s AI Copyright Bill Draws Platform Objections
Google, Meta and TikTok challenged parts of a proposed copyright revision covering news links and AI training. The draft is still under discussion and is not law.
What the proposal could change
The revision brings news compensation and AI copyright rules into one legislative process.
Payments for news use
An earlier draft described compensation for platforms that aggregate, republish or display news previews.
Training and disclosure
An earlier version described licensing or fair-use rules for training, plus disclosure requirements.
Why platforms object
Their concerns differ across links, AI training and intermediary protections.
| Company | Stated concern |
|---|---|
| Link payments, training licenses, collective licensing and safe harbor | |
| Meta | Links are not republication; payment duties could reduce publisher traffic and income |
| TikTok | AI disclosure should not apply when AI only assists a person’s work |
What readers and publishers face
The final rules could affect how news reaches readers and what material AI systems can use.
- News visibility in search and social products
- Publisher reach and bargaining position
- Copyrighted material available for AI training
- Final wording and scope of payment duties
- How licensing or fair use would work
- Whether platforms change their services
Rules Could Reshape News and AI
The proposal brings two contentious copyright questions into one legislative process: how platforms compensate news publishers when they surface their work, and whether AI developers need permission or a license to train models on copyrighted material. The result could affect publishers’ online reach, platform operations and the material available to AI systems trained for Indonesian-language use.
News publishers and technology companies describe the likely effects differently. At the hearing, the Press Council’s deputy chair, Totok Suryanto, said platforms take 80 percent of digital advertising and that traffic from platforms to news sites has fallen by up to 70 percent, according to the source reports. Those figures were presented by the Press Council; the hearing record described here does not establish them as independently verified measurements. Google, meanwhile, warns that charges for links could lead it to stop distributing local news through its services.
For readers, the practical stakes include whether news articles continue to appear in search and social products, and how any restrictions or licensing requirements affect AI tools. The draft’s scope and enforcement provisions could also influence publishers’ bargaining position. None of those outcomes is settled: they depend on the final law, how it is implemented, and whether any payment or licensing system is workable.
Products Worth Considering
Apple iPad Air 13-inch (M4), 128GB, Wi-Fi 7, Blue
LG UltraGear evo 27-inch 5K Mini-LED Gaming Monitor
Logitech MX Master 4 Ergonomic Wireless Mouse with Haptic Feedback, Graphite
As an affiliate, we earn on qualifying purchases.
Draft Links AI and News Rights
Reuters reported the draft’s AI provisions on July 17, citing a version that law ministry intellectual-property official Hermansyah Siregar confirmed was authentic. Siregar said it would explicitly recognize AI in Indonesian copyright law, while stressing that the text was not final. The version described by Reuters proposed rules for AI-assisted work, disclosure of AI use, imitation of a creator’s distinctive style, and use of copyrighted material for training under fair-use provisions or licensing agreements.
Under that July version, works generated entirely by AI would not qualify for protection, while AI-assisted works would need to meet a human-involvement threshold. The draft did not specify how much human input would be sufficient. It also proposed compensation for platforms that aggregate, republish or display news previews, including use of news for AI training, with payments routed through state-supervised collective management organisations. Possible sanctions described in the reporting included revoking local business permits. Those details come from an earlier draft, not a verified copy of the text now under discussion.
The proposal follows Presidential Regulation No. 32 of 2024, which requires digital platforms to support quality journalism. At the November hearing, the Press Council said compliance with that existing rule had been low, according to Aktual. The current copyright proposal would therefore add a new legislative framework to a policy area already covered by a presidential regulation.
“A levy on links would leave Google with “no choice” but to stop crawling, indexing and distributing Indonesian news.”
— Google representative Agung Pamungkas, as reported by Reuters
The full text now before lawmakers was not available in the source reporting, and the proposal has changed since the July version cited by Reuters. Reports refer to different versions, so details such as article numbers and the precise reach of payment obligations should be treated cautiously. The available material does not establish the final wording of the rules for AI training, disclosure, human contribution or news-link payments.
It is also unclear how lawmakers would define a qualifying AI-assisted work, which uses of copyrighted works would count as fair use, and how any collective licensing body would set and distribute payments. The companies’ warnings about access, traffic, investment and technical feasibility remain their stated assessments, not settled consequences. No date for a parliamentary vote has been announced.
Lawmakers Continue Bill Deliberations
Parliament’s Legislation Body and the government are continuing to deliberate the revision. A later draft or public release of the text would help clarify whether the provisions described in July remain in place and how lawmakers have responded to the platforms, publishers and press representatives heard on November 11.
No vote date has been announced, and the draft is not law. The next clear milestones are the publication or circulation of a more definitive text, further parliamentary discussion and any eventual vote. Until then, the scope of the proposed requirements—and whether platforms would change their services in response—remains open.
Key Questions
What is Indonesia’s proposed AI copyright bill?
It is a proposed revision of Law No. 28 of 2014 that would update copyright rules for AI and online platforms. Earlier reporting described provisions on AI-assisted works, disclosure, training data and payments related to news content, but the current text has not been fully confirmed.
Why are Google, Meta and TikTok objecting?
Google objects to payments for links and snippets, mandatory licensing for AI training, a single collective licensing body and changes to safe-harbor protections. Meta says sharing a link is not republication and argues payment duties could hurt publishers’ traffic and revenue. TikTok says disclosure should not be required when AI only assists a person’s work.
Would the bill require payment for every news link?
The July draft described by Reuters proposed compensation for certain platform uses of news, including aggregation and link previews. Google told lawmakers it understood the current draft to require payment for some indexing and display of news snippets. The current full text is unavailable, so the exact scope is not settled.
Is the bill already law?
No. It remains under deliberation by parliament’s Legislation Body and the government. No vote date has been announced.
What happens next?
Lawmakers are expected to continue deliberating the proposal. A newer draft, further hearings or a parliamentary vote could clarify the rules, but the available reporting gives no timetable for those steps.
Source: ThorstenMeyerAI.com
