(BSA): Recommendations for Indonesia's National AI Roadmap and Ethical Guidelines

Key Chapter Title List

1. BSA Resource Overview
2. AI Ethics Guidelines Should Be Voluntary
3. Definition Setting Aligned with International Understanding
4. Roles and Responsibilities in the AI Ecosystem
5. Risk-Based Approach
6. Self-Assessment and Incident Reporting
7. Consultation with Stakeholders
8. Conclusion

Document Introduction

This document is a formal policy recommendation submitted by the Business Software Alliance (BSA) to the Indonesian Ministry of Communication and Digitalization (KOMDIGI) on August 22, 2025. This recommendation responds to the Indonesian government's public consultation on the "National AI Roadmap White Paper (Draft)" and the "Artificial Intelligence Ethics Guidelines (Draft)." It aims to provide professional advice from the perspective of the global leading enterprise software industry to assist Indonesia in constructing its national AI strategy and governance framework. As a global industry association encompassing dozens of top technology companies such as Adobe, Microsoft, Amazon Web Services, IBM, and OpenAI, BSA has long been attentive to Indonesia's digital policy development and has previously engaged deeply on issues including personal data protection laws, electronic system operators, and national data centers.

The core issue of the report lies in balancing the incentives for AI innovation with risk control. BSA explicitly recommends that the Indonesian government clarify the non-binding and voluntary compliance nature of its "AI Ethics Guidelines" to avoid imposing unnecessary restrictions on technological development. The document emphasizes that if the government intends to make certain provisions legally binding, it must issue a clear declaration accompanied by multiple rounds of in-depth stakeholder consultations to ensure the clarity, feasibility, and alignment of the provisions with international best practices. This position sets the tone for all subsequent BSA recommendations, advocating for a flexible, risk-based, and internationally aligned governance model.

At the level of specific recommendations, the report highlights three main pillars. First is the international alignment of definitions and frameworks. BSA strongly recommends that Indonesia adopt the updated AI definition from the Organisation for Economic Co-operation and Development (OECD) and refer to the risk management frameworks of the OECD, the Association of Southeast Asian Nations (ASEAN), and the U.S. National Institute of Standards and Technology (NIST) for role definition (e.g., developer, integrator, deployer) and responsibility allocation. This international coordination is seen as key to reducing regulatory conflicts, promoting cross-border cooperation, and establishing global safety benchmarks. Second, the report advocates for a strictly risk-based, tiered governance approach. Governance efforts should focus on "high-risk" AI application scenarios that could cause significant harm to individuals or society (such as decisions concerning housing, credit, employment), while maintaining a relaxed approach for low-risk applications to unleash their economic benefits. Third, regarding implementation mechanisms, BSA recommends that any self-assessment requirements should be limited to clearly defined high-risk use cases and undertaken by the most appropriate party in the supply chain (e.g., developers assessing training data, deployers assessing usage scenarios). The report particularly opposes establishing an independent, duplicate reporting framework for AI incidents, arguing that existing cybersecurity and critical infrastructure incident reporting channels should be utilized to avoid imposing unnecessary compliance burdens on businesses.

Finally, BSA offers recommendations on the policy-making process itself. Citing the complexities encountered during the legislative process of the EU's "AI Act" as an example, the report warns that overly rushed timelines could lead to flawed frameworks or impractical recommendations. Therefore, BSA recommends that the Indonesian government extend the consultation deadline by one month to September 19 and set clear milestones for the implementation of the ethics guidelines. This aims to allow space for more thorough dialogue with experts, industry, and civil society, thereby developing balanced and effective guidelines that align with both global best practices and Indonesia's unique national conditions and priorities. The entire document reflects the industry's urgent desire to collaborate with policymakers to jointly shape an AI governance environment that fosters innovation while responsibly managing risks.