Investigation Report: Implementation of the "Act on Promotion of Research and Development and Utilization of Artificial Intelligence Related Technologies" and the "AI Basic Plan"
Key Points
Official Framework: The "Act on Promotion of Research and Development and Utilization of Artificial Intelligence Related Technologies" (commonly referred to as the AI Promotion Act or AI Act) was promulgated in June 2025. Following this, the "AI Basic Plan" was Cabinet-decided in December 2025, with full-scale support measures and implementation frameworks commencing in January 2026 [cite: 1, 2, 3]. Policy Stance: Japan maintains a "soft law" approach, prioritizing innovation and voluntary governance over strict penalties, distinguishing it from the European Union's "hard law" regulations. The focus is on balancing "promotion of innovation" with "risk management" [cite: 4, 5]. SME Impact: Small and Medium-sized Enterprises (SMEs) are classified as "Utilization Business Operators." While there are no criminal penalties for non-compliance, businesses are obligated to cooperate with government policies and strive for transparency. The government has allocated significant budgets for subsidies to support AI introduction and security risk assessments starting in 2026 [cite: 6, 7, 8]. Strategic Goals: The government aims to make Japan "the world's most AI-friendly country for development and utilization," targeting a reversal of the current lag in AI adoption compared to the US and China [cite: 2, 9].
1. Confirmation of Official Information
This section details the statutory framework and official announcements that form the basis of the new regulations and support systems effective from January 2026.
1.1 Official Announcements and Legislative Framework
The policy in question is grounded in the "Act on Promotion of Research and Development and Utilization of Artificial Intelligence Related Technologies" (Act No. 53 of 2025).
Official Name: Act on Promotion of Research and Development and Utilization of Artificial Intelligence Related Technologies (人工知能関連技術の研究開発及び活用の推進に関する法律) [cite: 1, 10]. Common Names: AI Promotion Act (AI推進法), AI Act (AI法), AI New Law (AI新法) [cite: 1, 7]. Relevant Ministries: Cabinet Office (leading the AI Strategy Council), Ministry of Economy, Trade and Industry (METI), Ministry of Internal Affairs and Communications (MIC) [cite: 4, 11].1.2 Timeline of Enactment and Enforcement
Promulgation Date: June 4, 2025 [cite: 1]. Partial Enforcement: June 4, 2025 (General provisions). Full Enforcement: The provisions regarding the "AI Strategy Council" and the "AI Basic Plan" (Chapters 3 and 4) were enforced within three months of promulgation (September 2025) [cite: 1, 5]. Operational Commencement (User Query Context): While the law itself was legally effective by late 2025, the "AI Basic Plan"—which dictates the actual budget, subsidies, and specific guidelines—was Cabinet-decided on December 23, 2025 [cite: 2, 3]. Consequently, the practical application of support measures and the full-scale operation of the environment for SMEs effectively began in January 2026 [cite: 12, 13].1.3 Scope of Target Businesses
The law broadly categorizes stakeholders into three groups. SMEs generally fall under "Utilization Business Operators."
1 Research and Development Agencies: Entities developing AI technologies (e.g., foundation model developers).
2 Utilization Business Operators (活用事業者): Entities using AI for business purposes. This includes almost all SMEs that introduce AI tools (like ChatGPT, SaaS AI, or custom AI systems) into their workflows [cite: 5, 10].
3 The Public (Citizens): General users who utilize AI services [cite: 10].
2. Background and History
Understanding why this policy was introduced is crucial for SMEs to align their business strategies with national interests.
2.1 Rationale for Policy Formulation
The primary driver is the "sense of crisis" regarding Japan's international standing in AI adoption.
Lag in Adoption: As of 2024, the utilization rate of generative AI in Japanese companies was approximately 47%, significantly lower than the ~85% seen in the US and China. The government explicitly states that "not using AI is the biggest risk" [cite: 14, 15]. Economic Stagnation: With a shrinking population and labor shortage, AI is viewed as a critical "crisis management investment" and "growth investment" to maintain national productivity [cite: 3].2.2 International Trends and Japan's Position
Japan's approach is strategically positioned between the US and the EU.
EU (Hard Law): The "EU AI Act" imposes strict prohibitions and heavy fines for high-risk AI, prioritizing fundamental rights [cite: 4, 16]. US (Innovation First): Focuses on security risks (national defense) while promoting private sector innovation [cite: 16]. Japan (Soft Law/Agile Governance): Japan aims to be "the most AI-friendly country" by avoiding rigid legal restrictions that might stifle innovation. Instead, it relies on "binding soft law" where government guidelines and "duty to cooperate" drive compliance [cite: 4, 5]. This aligns with the "Hiroshima AI Process" led by Japan in the G7 to promote safe, secure, and trustworthy AI [cite: 3, 16].2.3 Changes from Past Policies
Previously, AI governance relied solely on non-binding guidelines (e.g., "AI Principles"). The AI Promotion Act elevates this to a statutory basis.
From "Guidelines" to "Law": While still "soft," the establishment of the AI Strategy Council and the legal obligation to formulate a Basic Plan ensures a permanent, funded, and coordinated government effort, rather than ad-hoc measures [cite: 1, 7]. Centralized Command: The establishment of the "AI Strategy Council" (chaired by the Prime Minister) unifies policies that were previously scattered across METI, MIC, and the Digital Agency [cite: 7, 17].
3. Specific Content (Point-by-Point Commentary)
This section breaks down the legal obligations and the content of the "AI Basic Plan" relevant to business owners.
3.1 Key Provisions of the AI Promotion Act
The law consists of 28 articles focused on "Promotion" rather than "Regulation" [cite: 4, 7].
1 Basic Philosophy (Article 3):
* AI development and use must ensure "transparency," "fairness," and "protection of rights."
* It must contribute to "human-centered" society [cite: 5, 10].
2 Responsibilities of Utilization Business Operators (Article 7):
* SMEs using AI are required to "strive to use AI properly" in accordance with the basic philosophy.
* There is a duty to cooperate with measures implemented by the national and local governments (e.g., surveys, safety reporting) [cite: 7, 10].
3 Establishment of the AI Strategy Council (Article 19):
* A "control tower" under the Cabinet to lead policy. This council determines the "AI Basic Plan" [cite: 5, 7].
3.2 The "AI Basic Plan" (Decided Dec 2025)
The Basic Plan, which dictates the actual support measures starting Jan 2026, is built on four pillars [cite: 3, 8, 18]:
1 Using AI (AIを使う): Drastic expansion of AI use in government and SMEs.
2 Creating AI (AIを創る): Strengthening domestic development capacity (computing power, foundation models).
3 Enhancing Reliability (信頼性を高める): Establishing safety evaluation methods (via the AI Safety Institute).
4 Collaborating with AI (AIと協働する): Reskilling and labor market reforms.
3.3 Penalties and Risks
No Criminal Penalties: Unlike the EU AI Act, the Japanese law does not impose fines or imprisonment for violation of the act itself [cite: 4, 19]. "Name and Shame" Risk: The government has the authority to investigate and provide guidance. If a business operates AI inappropriately (e.g., violating human rights or causing accidents) and refuses to follow government guidance, the company name may be publicly announced. This poses a significant reputational risk [cite: 7, 17]. Commercial Risk: Non-compliance with the "AI Basic Plan" guidelines may result in exclusion from government procurement or ineligibility for the subsidies described below.
4. Impact on SMEs and Required Actions
This is the most critical section for the target audience (SME owners).
4.1 Target Audience
Any SME that utilizes AI technology is a "Utilization Business Operator."
*Examples:* A manufacturing firm using AI for visual inspection; a retail shop using ChatGPT for marketing copy; a clinic using AI for appointment scheduling [cite: 5, 20].4.2 Required Actions (What to do)
Based on the "AI Basic Plan" and the "AI Business Operator Guidelines" (updated alongside the law), SMEs should take the following steps:
1 Risk Assessment:
* Before introducing AI, assess risks related to "Copyright Infringement," "Personal Information Leakage," and "Hallucinations" (false information) [cite: 5, 21].
* *Action:* Create an internal checklist or use government-provided assessment tools.
2 Security Measures:
* Implement measures to prevent data leakage (e.g., opting out of training data usage in cloud AI settings).
* *Conditionality:* The news indicates that "security measures" are a condition for receiving subsidies.
3 Literacy Education:
* Employees must be trained not just on *how* to use AI, but on the *ethics* and *risks* (e.g., not inputting confidential customer data into public AI models) [cite: 9, 21].
4.3 Support Measures (Subsidies & Resources)
The government has prepared "Crisis Management Investment" and "Growth Investment" budgets [cite: 3].
Subsidies for Labor Saving (省力化補助金): Financial support for introducing AI systems (software/hardware) that improve productivity. This is likely part of the "SME Productivity Revolution Promotion Project" (budget ~340 billion JPY) [cite: 6, 22]. Risk Assessment Support: The government (via agencies like IPA or the AI Safety Institute) provides tools and potentially consulting subsidies to help SMEs evaluate AI safety [cite: 3, 8]. Reskilling Support: Grants for employee training programs focused on AI utilization [cite: 8].4.4 Estimated Costs and Resources
Financial Cost: Direct costs for compliance are low (no expensive certification fees mandated yet). The main cost is the *time* spent on governance and the purchase of AI tools (which may be subsidized by 1/2 to 2/3) [cite: 22]. Human Resources: SMEs need to appoint a "person in charge of AI" (does not need to be a new hire, but an existing manager) to oversee guidelines and security [cite: 21].
5. Future Outlook
5.1 Upcoming Policy Developments
Annual Updates: The AI Basic Plan is designed to be "Agile," meaning it will be reviewed and potentially revised every year. SMEs must stay updated on annual changes [cite: 9, 11]. Guidelines for "Physical AI": Future regulations will likely focus on "Physical AI" (robots, autonomous driving, drones) where safety risks are physical. Manufacturing SMEs should watch this closely [cite: 11, 23].5.2 Possibility of Hard Law
While the current law is "soft," the government reserves the right to introduce legal regulations (Hard Law) if voluntary efforts are deemed insufficient or if high-risk incidents occur frequently. The "soft law" approach is a trial period for the private sector to demonstrate self-governance [cite: 24, 25].
5.3 Preparation for the Future
SMEs should view AI not just as a tool but as a strategic asset.
Data Preparation: The Basic Plan emphasizes "Data." SMEs should start digitizing and organizing their internal data now to be ready for AI integration [cite: 9, 26]. "Human-Centric" Operations: As AI takes over routine tasks, the value of human judgment increases. Defining the "Human-AI Division of Labor" is the next management challenge [cite: 8, 21].
6. Conclusion (For the Blog Post)
The full enforcement of the AI Promotion Act in 2026 signals the Japanese government's serious commitment to AI. For SMEs, this is not a regulatory headwind but a tailwind of support. By utilizing the "Risk Assessment Support" and "Subsidies" provided under the AI Basic Plan, SMEs can introduce cutting-edge technology safely and at a reduced cost. The key is to move from "vague anxiety" to "managed risk taking."
Reference Links:
Cabinet Office: AI Strategy (Official Source) e-Gov: AI Promotion Act Full Text [cite: 10] METI: AI Business Operator Guidelines (Inferred context)Sources:
