Japan's Financial Services Agency has no AI-specific binding rule for banks: its position is the AI Discussion Paper, published as Version 1.0 on 4 March 2025 and updated as Version 1.1 on 3 March 2026, which keeps a technology-neutral stance (existing laws apply regardless of whether AI is used) and encourages sound adoption. The updated paper adds findings from the FSA AI Public-Private Forum (June to December 2025) on customer-facing AI and AI agents. On 22 May 2026 the FSA and the Bank of Japan jointly requested financial institutions to take nine short-term measures, over roughly one month, against the cyber threat of frontier AI. The discussion papers are non-binding and the request builds on existing cybersecurity supervision.
| Full name | Financial Services Agency of Japan (with the Bank of Japan on frontier AI cyber measures) |
| Role | Integrated financial supervisor whose AI position is a dialogue-based discussion paper and, in 2026, a joint request with the Bank of Japan on frontier-AI cyber threats |
| Force on banks | Supervisory guidance |
| Applies to | Financial institutions in Japan, including banks. The AI Discussion Paper is non-binding and states that existing laws apply regardless of whether AI is used; the May 2026 request to financial institutions is a supervisory request addressed to heads of relevant firms |
| Key document | AI Discussion Paper (Version 1.1), 3 March 2026, updating Version 1.0 of 4 March 2025 |
| Latest move | 22 May 2026: the FSA and the Bank of Japan jointly requested financial institutions to implement nine short-term responses to frontier AI cyber threats, expected over roughly one month. |
| Documents tracked | 3 · all documents → |
The FSA's discussion paper is built on a survey of 130 firms in autumn 2024 and rejects a rules-first approach: legal regulation should be limited to areas where self-regulation cannot be expected, so principles and guidelines would be revised before any legal change. Version 1.1 records that customer-facing generative AI had begun in limited-scope deployments, groups risk mitigation for such services into four perspectives, and gives the FSA's reading of securities-sector questions on outsourcing data, access to corporate-related information and solicitation.
Frontier AI changed the agenda in 2026. After an FSA-hosted public-private meeting on 24 April 2026 and a first practitioner working group on 14 May, the FSA and the Bank of Japan issued a request on 22 May 2026 asking institutions, under direct management involvement, to prioritise critical systems, secure patching capacity, make patching risk-based and prepare for proactive shutdown of services. The request follows the national government's Project YATA-Shield package of 18 May 2026.
What has the Japan FSA actually published on AI?
| Date | Document | Status |
|---|---|---|
| May 22, 2026 | Japan FSA/BOJ frontier AI request — Request regarding 'Short-term Responses by Financial Institutions in Light of Threat Changes from Frontier AI' (「フロンティアAIによる脅威変化を踏まえた金融機関等の短期的な対応」に係る要請について) | In force |
| Mar 3, 2026 | Japan FSA AI Discussion Paper v1.1 — AI Discussion Paper (Version 1.1): Preliminary Discussion Points for Promoting the Sound Utilization of AI in the Financial Sector | Final |
| Mar 4, 2025 | Japan FSA AI Discussion Paper v1.0 — AI Discussion Paper (Version 1.0): Preliminary Discussion Points for Promoting the Sound Utilization of AI in the Financial Sector | Superseded |
| Date | Type | Document / event |
|---|---|---|
| May 22, 2026 | Letter | Japan FSA/BOJ frontier AI request — Request regarding 'Short-term Responses by Financial Institutions in Light of Threat Changes from Frontier AI' (「フロンティアAIによる脅威変化を踏まえた金融機関等の短期的な対応」に係る要請について). On 22 May 2026 the Japan Financial Services Agency and the Bank of Japan jointly requested financial institutions to take short-term measures against the changed cyber threat posed by frontier AI, which can discover vulnerabilities and generate attack code far faster than before. source ↗ |
| Apr 24, 2026 | Milestone | FSA public-private meeting on strengthening financial-sector cybersecurity against AI threats. The FSA held a public-private meeting on 24 April 2026; a first practitioner-level working group followed on 14 May 2026 and produced the short-term response document attached to the 22 May 2026 request. |
| Mar 3, 2026 | Report | Japan FSA AI Discussion Paper v1.1 — AI Discussion Paper (Version 1.1): Preliminary Discussion Points for Promoting the Sound Utilization of AI in the Financial Sector. The Japan Financial Services Agency released Version 1.1 of its AI Discussion Paper on 3 March 2026, updating the March 2025 Version 1.0 with insights from the FSA AI Public-Private Forum held from June to December 2025. source ↗ |
| Jun 2025 | Milestone | FSA AI Public-Private Forum begins (June to December 2025). The forum shared knowledge on AI use, risk-management and governance examples, and regulatory clarification needs; its findings led to Version 1.1 of the AI Discussion Paper. |
| Apr 10, 2025 | Milestone | English version of the AI Discussion Paper (Version 1.0) published. The FSA published the English version of the AI Discussion Paper on 10 April 2025, five weeks after the Japanese original of 4 March 2025. |
| Mar 4, 2025 | Report | Japan FSA AI Discussion Paper v1.0 — AI Discussion Paper (Version 1.0): Preliminary Discussion Points for Promoting the Sound Utilization of AI in the Financial Sector. The Japan Financial Services Agency published its AI Discussion Paper (Version 1.0) in Japanese on 4 March 2025, with an English version on 10 April 2025, as an initial set of discussion points for the sound use of AI by financial institutions. source ↗ |
- Further revisions of the AI Discussion Paper and any FSA guideline changes, which the FSA says come before any legal response.
- Follow-up on the 22 May 2026 frontier-AI request, including any review of the short-term measures as AI developments change.
- Regulatory clarifications from case-specific dialogue with industry bodies on AI use in securities and other sectors.
Does Japan have an AI regulation for banks?
There is no AI-specific binding rule for banks. The FSA's AI Discussion Paper states a technology-neutral stance: existing laws and supervisory guidelines apply whether or not AI is used, and the FSA would revise principles and guidelines before considering legal changes.
What is the latest version of the Japan FSA AI Discussion Paper?
Version 1.1, published on 3 March 2026, which updates Version 1.0 of 4 March 2025 using findings from the FSA AI Public-Private Forum held from June to December 2025.
What did the FSA and the Bank of Japan ask banks to do about frontier AI?
On 22 May 2026 they requested nine short-term measures, expected over roughly one month, including management-level prioritisation, risk-based patching, vendor contract checks, stronger non-patch defences and readiness to shut down priority systems.
Follow every move these regulators make
when one of these regulators moves, the next morning's brief says so · six sourced stories · 7 am ET · free
plus every tracker, bank and agent page update, the morning after · leave any morning