PHL Supreme Court adopts AI rules – but humans still decide

By Joseph Bernard A. Marzan

MANILA — The Supreme Court En Banc has recently adopted a governance framework regulating the use of artificial intelligence tools in the Philippine judiciary, emphasizing human oversight, transparency, and accountability in the use of emerging technologies.

The Supreme Court issued a press release on Friday, March 20, that the En Banc, the court’s highest body, adopted the Governance Framework on the Use of Human-Centered Augmented Intelligence in the Philippine Judiciary in a resolution in A.M. No. 25-11-28-SC dated Feb. 18, 2026.

The framework is anchored on three ethical principles: fairness, accountability, and transparency, aiming to ensure the responsible use of artificial intelligence in the judicial system.

The court said the rapid progress in artificial intelligence has affected nearly all aspects of society, including the justice sector, legal practice, and court operations.

However, it warned that AI systems used without proper human oversight may cause harm, including “amplification of preexisting biases, inequalities, discrimination, and widening of the digital divide; propagation of misinformation, disinformation, and malinformation; and the acceleration of the climate emergency.”

The framework was developed by a working group chaired by Senior Associate Justice Marvic M.V.F. Leonen, with Associate Justices Ramon Paul L. Hernando and Rodil V. Zalameda as vice chairpersons, in consultation with members of the judiciary, subject matter experts, lawyers, and the academe.

It also drew on international standards, including guidance from the Association of Southeast Asian Nations and the United Nations Educational, Scientific and Cultural Organization.

The Philippines is among the first countries in Southeast Asia to adopt a formal AI governance framework specifically for its judiciary, reflecting a growing global trend among courts to establish guidelines for the responsible use of AI in legal proceedings and court administration.

A key feature of the framework is the use of the term “human-centered augmented intelligence,” which emphasizes that artificial intelligence must remain centered on humans and should support, and not replace, human reason and judgment.

The resolution states that human-centered augmented intelligence tools should be used “as a support or augmentation tool that complements and empowers human cognitive skills, rather than completely supplanting or overriding human judgment.”

The framework also highlights human values such as the rule of law, fundamental freedoms, dignity and autonomy, privacy and data protection, fairness, nondiscrimination, and social justice, and includes environmental responsibility and sustainability in the use of AI technologies.

The court also stressed that under no circumstances should AI tools serve as the sole basis for adjudicatory decisions, and that legal reasoning and final conclusions must still be made by human decision-makers.

The framework requires disclosure when AI tools are used in judicial work, including in tasks such as legal research, summarization, translation, proofreading, document processing, and voice-to-text transcription.

Users must disclose the AI tool used, its version, the purpose for using it, the extent of AI involvement, and the level of human oversight, and must declare that they remain responsible for the output in line with the principle of accountability.

Disclosure is mandatory for members of the judiciary, court officials, and employees when AI tools are used in preparing court-issued documents, including voice-to-text transcription, translation, automated compilation of authorities or citations, legal research, document summarization, automated document processing such as optical character recognition, copy editing or proofreading, and data redaction or sanitation.

The resolution states that responsibility ultimately lies with the designer, developer, or user of an AI tool, and that users remain personally responsible for the tool’s output and its consequences.

The court emphasized that AI tools must not reduce human control, management, or supervision, and cannot be used if they could harm stakeholders, violate rights, or undermine the rule of law.

AI tools or their outputs must not be the sole basis for any adjudicatory decision, as human decision-makers remain responsible for independent legal reasoning and final judgments.

The framework also requires risk assessments before any AI tool is used, safeguards for privacy and data protection, training programs to address algorithmic and automation bias, and continued monitoring, auditing, and cybersecurity measures to ensure responsible use of the technology.

The new AI governance framework applies to justices and judges of all court levels, court officials and employees, court users including lawyers and litigants, and vendors or third-party contractors involved in designing, developing, or using AI tools for the judiciary.

No AI tool may be used unless authorized by the Supreme Court En Banc, and implementation will be phased, starting with pilot testing before a gradual rollout.

The court will also establish a permanent Committee on Human-Centered Augmented Intelligence to guide the design, development, and ethical use of AI tools in the judiciary.

The framework supports the Supreme Court’s Strategic Plan for Judicial Innovations 2022–2027, which aims to build a technology-driven judiciary that is transparent, accountable, and accessible, while ensuring that artificial intelligence is used ethically and responsibly in the administration of justice.