IP Concerns In Machine-Generated Translations Of Barangay Ordinances.
1. Introduction: Machine-Generated Translations and IP
Barangay ordinances are local laws in the Philippines governing communities. With advances in AI and machine translation, platforms may automatically translate these ordinances from Filipino to English, or vice versa, for accessibility.
IP concerns arise because:
Machine translation involves copyrightable source content (the original ordinances may be subject to copyright, depending on authorship and government ownership).
Machine-generated translations raise questions of authorship and ownership of derivative works.
Use of third-party machine translation software may involve licensing or copyright restrictions.
Misuse of translations can result in liability for misrepresentation or plagiarism.
Key IP issues:
Copyright in original ordinances
Derivative work rights in translations
Authorship of AI-generated content
License compliance for translation tools
Database and compilation rights for organized ordinances
2. Intellectual Property Concerns
(A) Copyright in Barangay Ordinances
Generally, Philippine law under the Copyright Law (Republic Act No. 8293) protects literary works, including text of legal documents.
However, government works are often public domain, but this depends on the jurisdiction:
Original text drafted by a private legal consultant or lawyer may be copyrighted.
Text authored by government officials in the course of official duties may be in the public domain.
Risk: AI translation platforms may treat all ordinances as free content, potentially infringing private copyright if not verified.
(B) Machine-Generated Translations as Derivative Works
A translation is considered a derivative work, protected under copyright law if it involves human creativity.
Key legal question: Who owns a machine-generated translation?
If fully automated, most jurisdictions do not grant copyright to AI (see U.S. Copyright Office policy).
Human supervision or editing may create copyrightable derivative work.
Risk: Republishing AI translations without authorization could infringe derivative work rights.
(C) Software Licensing Issues
Using machine translation software (e.g., Google Translate, OpenAI APIs) involves license agreements.
Violating terms (e.g., using translations commercially without a license) can trigger contractual and copyright liability.
(D) Data and Database Rights
Compiled ordinances (organized by year, topic, or region) may be protected as databases.
Unauthorized copying or extraction of large amounts of data can infringe database rights.
(E) Moral and Attribution Rights
In jurisdictions like the Philippines, authors have moral rights (RA 8293, Section 88) even if the work is in the public domain.
Translating and publishing without attribution may violate these rights.
3. Case Laws Relevant to Machine-Generated Translations
Here are more than five detailed cases illustrating the IP issues:
Case 1: Feist Publications, Inc. v. Rural Telephone Service Co., Inc. (1991, U.S.)
Facts
Feist published a directory using data from Rural Telephone Service without permission.
Rural claimed copyright infringement on its compiled telephone listings.
Issue
Whether facts and compilations without originality are protected by copyright.
Judgment
Supreme Court held that facts are not copyrightable, only original selection or arrangement is.
Compilations with minimal originality cannot be copyrighted.
Significance
For Barangay ordinances, raw legislative text may not be copyrightable, but organized compilations may be protected.
Machine translation must consider originality of compiled content.
Case 2: Microdecisions, Inc. v. Skinner (2004, Florida)
Facts
Government-created GIS maps were restricted for commercial use.
Private company challenged copyright claim.
Judgment
Court ruled that government-produced public records cannot be copyrighted.
Significance
Similarly, Barangay ordinances authored by government officials are likely public domain, so AI translation is generally permissible.
However, translations introducing human creativity may create new IP claims.
Case 3: Naruto v. Slater (2016, U.S.)
Facts
A monkey took a selfie; Slater claimed copyright over the photo.
Issue
Whether non-human authors can hold copyright.
Judgment
Court held that non-human authors cannot hold copyright.
Significance
Applies to machine-generated translations:
Fully AI-generated translations may not have copyright protection unless there’s human creative input.
Case 4: Jacobsen v. Katzer (2008, U.S.)
Facts
Open-source software was used commercially without complying with license terms.
Judgment
Violation of open-source licenses constitutes copyright infringement.
Significance
If a machine translation platform is used with restrictive software licenses, users must comply with terms.
Unauthorized redistribution of translations can lead to legal liability.
Case 5: Bilski v. Kappos (2010, U.S.)
Facts
Patent on a business method was challenged for being abstract.
Judgment
U.S. Supreme Court confirmed that abstract ideas cannot be patented.
Significance
AI-generated translation algorithms may not be patentable unless they meet patent criteria.
Protecting AI translation software/processes requires careful patent drafting.
Case 6: Authors Guild v. Google (2015, U.S.)
Facts
Google scanned books and provided snippets for search purposes.
Judgment
Court held that use of copyrighted work for transformative purposes (search indexing) is fair use.
Significance
AI translations may be considered transformative works if used for accessibility or research, potentially falling under fair use, but commercial use may require licensing.
Case 7: U.S. Copyright Office Guidance – AI Authorship (2022)
Facts
AI-generated content submitted for copyright registration.
Guidance
Works created solely by AI without human authorship are not copyrightable.
Human involvement is required to claim copyright.
Significance
Machine-generated translations of ordinances cannot be copyrighted unless human edits or contributions are substantial.
This affects commercial exploitation of AI translations.
4. Practical IP Risk Assessment
| IP Concern | Risk in Machine Translation of Ordinances | Mitigation |
|---|---|---|
| Original Ordinances | Usually low if public domain; higher if authored by private lawyers | Verify authorship, consult legal counsel |
| Translation as derivative work | Human-edited AI translations may create new copyright | Ensure human authorship or proper licensing |
| Software license | High if using proprietary translation software | Comply with licenses, use open-source alternatives carefully |
| Database rights | Medium if compiling large ordinance datasets | Attribute sources, avoid mass extraction without permission |
| Moral rights | Medium if human authors involved | Proper attribution, acknowledgment |
5. Legal Strategies
Verify original authorship – Determine if the ordinance is public domain or protected by copyright.
Use licensed translation tools – Ensure commercial AI usage complies with terms.
Maintain attribution – If a human editor contributed, properly credit them.
Document human contributions – To claim copyright in derivative translations.
Use fair use/exceptions carefully – For research or accessibility purposes.
6. Conclusion
Machine-generated translations of Barangay ordinances raise multiple IP concerns:
Ownership of derivative translations
Licensing of AI software
Database rights for compiled ordinances
Moral rights and attribution
Key takeaways from case law:
AI alone cannot hold copyright (Naruto v. Slater, USCO Guidance)
Government works are often public domain (Microdecisions v. Skinner)
Derivative works with human creative input may have protection
License compliance is critical (Jacobsen v. Katzer)
Transformative use may qualify as fair use (Authors Guild v. Google)
RELATED Blog
comments