Hey there,

In April, the Japanese Cabinet approved a bill that repositioned Japan as one of the most permissive AI development environments in the developed world.

The bill is an amendment to Japan's Act on the Protection of Personal Information (APPI). It creates new pathways for companies to use personal data to train AI models and build AI products, without requiring individual consent in many cases. Where Europe's GDPR assumes data sharing is risky until proven safe, Japan's revised framework assumes data use for research and AI development is acceptable unless proven harmful.

Prime Minister Takaichi has been explicit about the goal: "We will make Japan the world's best country to develop and use AI." The Minister for Digital Transformation has said Japan aims to be "the easiest place in the world to develop AI apps."

For foreign professionals, this matters on two levels. First, it is one of the clearest signals yet that Japan's AI job market is about to expand significantly, and that companies hiring for AI and data roles have a new regulatory tailwind behind them. Second, this is your personal data that the new rules apply to, and it is worth understanding what changed and what rights you retain.

This week: the full picture on Japan's APPI amendment, what it means for careers in AI and data, and how to position yourself for the hiring wave that follows.

Japan's push to become the world's leading AI development environment is not just a policy statement. It is a labour market signal, and if you are an AI, ML, or data engineer, it has direct implications for your visa options and negotiating position.

Which visa category covers AI roles:

Most AI and data engineering roles in Japan fall under the Engineer category of the Engineer / Specialist in Humanities / International Services visa. This covers software development, machine learning engineering, data science, and research engineering roles at private companies. AI product management, depending on the specific duties, can fall under either the engineer or the humanities/business subcategory. Confirming the correct category with your sponsoring employer before filing is standard practice.

Why the AI boom strengthens your HSP case:

The Highly Skilled Professional (HSP) visa is points-based, and AI engineering roles are unusually well-positioned to qualify. The combination of STEM degrees (20 to 30 points for Master's and Doctoral qualifications), strong salaries in the ¥7M to ¥15M range typical of ML and senior data engineering roles (30 to 40 salary points), and work experience in a high-demand field can put many candidates at or above the 80-point threshold for one-year PR eligibility.

The AI hiring boom adds one more lever: companies are competing harder for AI talent than for almost any other engineering profile. That competition translates into higher initial offers and more salary negotiation room. An offer that gets you from 30 to 40 salary points is worth specifically asking for, and an AI role gives you more leverage to do it than most engineering positions.

The salary thresholds to know:

Senior ML engineers and AI researchers at Japan's leading tech companies are regularly receiving offers in the ¥8M to ¥15M range. At that salary level combined with a standard STEM background, most candidates are already at 70 to 80+ points before language ability is factored in.

The privacy engineering opportunity:

The APPI amendment is creating a new role type in Japan that barely existed here two years ago: privacy engineer. Companies now need engineers who can build pseudonymization pipelines, purpose-limitation logging, consent management systems, and audit trail infrastructure at production scale. These roles fall squarely within the engineer visa category and are increasingly well-compensated given how few people have the combination of data engineering and compliance knowledge required.

One practical step:

If you are evaluating AI roles in Japan and have not run your HSP point total recently, now is the time. The AI hiring wave means salary offers are moving higher. An offer that clears ¥8M or ¥10M changes your PR timeline materially. Run the numbers before you negotiate, not after.

Japan's shift toward being the most permissive AI development environment in the G7 is going to move the job market. Here is how to make sure you are talking about it the right way in interviews.

The framing that lands:

Japanese companies building AI products are no longer primarily worried about whether the regulatory environment will let them ship. The constraint has shifted from "can we legally do this?" to "do we have the talent to do this?" That is where you come in, and it changes what you should be emphasizing.

In interviews at AI-adjacent companies, frame your experience around what you can build and ship, not just what you know. The technical depth matters, but so does your ability to move from model to product in a real regulatory and data environment. Japan's amended APPI gives companies more data access than they had before, but they still need engineers who know how to handle personal data responsibly, build systems with privacy-by-design principles, and work within the compliance safeguards the new rules require.

A script for the AI-era Japan interview:

This signals three things: you have done your research, you understand the regulatory context (which Japanese companies value highly), and you are framing Japan as an opportunity rather than a second-tier market. That last point lands better than it sounds. Many foreign engineers come to Japan with a "settling" framing. Showing genuine excitement about Japan's AI regulatory environment reads as authentic because it is a real, differentiated thing right now.

The roles that benefit most directly:

The APPI amendment creates the most immediate opportunity for engineers working at the intersection of data and product:

At international companies:

For engineers at foreign-affiliated firms with operations in Japan, the APPI amendment creates specific opportunities around cross-border data transfer. Companies training global models can now include Japanese user data under specific conditions. Engineers who understand both the technical and compliance dimensions of this, including pseudonymization, transfer risk assessments, and audit trail requirements, are in a strong position.

🇯🇵 今日の面接フレーズ (Today's Interview Phrase)

個人情報保護法の改正により、AIの開発環境が大きく変わると考えています。 Kojin jōhō hogohō no kaisei ni yori, AI no kaihatsu kankyō ga ōkiku kawaru to kangaete imasu. "I believe the APPI amendment will significantly change the environment for AI development."

Use this when opening a discussion about Japan's AI regulatory shift in an interview. It signals that you are tracking policy changes, not just technical trends. That distinction stands out in Japanese professional contexts, where demonstrating knowledge of the regulatory environment shows genuine engagement with working in Japan rather than just the technology.

Common mistake: Treating Japan's AI regulatory shift as purely a data compliance story. The companies that will benefit most from the APPI amendment are not the ones building compliance systems. They are the ones building AI products that were previously constrained by consent requirements. Position yourself as someone who can build those products, with compliance as a design dimension rather than the point.

The best person for your next role might not live near your office — or even in the same country.

And more companies are realizing they don’t need to open entities everywhere just to hire great talent globally.

Instead, teams are using EOR to hire internationally faster, stay compliant, and avoid the operational headache of setting up local infrastructure before they’re ready.

That shift is changing how companies think about growth, hiring, and expansion altogether.

Oyster’s EOR helps companies hire, pay, and support employees in 180+ countries while Oyster handles payroll, compliance, taxes, and local employment requirements.

Talk to experts today

Japan's AI market was already growing before April's amendment. The regulatory change accelerates an existing trend and removes one of the main structural constraints on domestic AI development.

The gap that's closing:

Japan's AI development sector has historically been hampered by two things: a shortage of training data (due to strict consent requirements) and a shortage of AI talent (due to limited appeal to international engineers compared to Silicon Valley or London). The APPI amendment addresses the first directly. The broader positioning of Japan as an AI-friendly jurisdiction addresses the second indirectly. More companies building more things means more demand for talent.

The sectors moving fastest:

The new data rules have the most immediate impact on industries that already hold large amounts of personal data and have the most to gain from AI.

HR technology: payroll data, employment records, social insurance filings were previously consent-limited for AI analysis and are now significantly more permissive. SmartHR, freee, and similar companies are directly positioned to build new AI features on top of the data they already hold.

Financial services: transaction data, credit data, and insurance records. The amended rules accelerate AI development in fraud detection, credit scoring, and financial planning tools.

Healthcare and pharma: patient data remains separately regulated under the Act on the Protection of Medical-Related Information, but adjacent health data flows benefit from the amended framework.

Retail and e-commerce: purchase history, user behavior data, and public-facing personal information. Companies like Rakuten, Mercari, and ZOZO have vast datasets that become more actionable for AI development.

The international company angle:

The APPI amendment creates a specific strategic advantage for Japan offices of global tech companies. Engineers at Google, Amazon, Microsoft, Stripe, and similar firms with Tokyo operations are now working in a jurisdiction that allows AI development patterns that are not available in EU markets. This is a meaningful career differentiator. The work you can do in Japan on AI products is different from the work you can do in Germany or France, and companies are aware of it.

The talent demand picture:

Japan's IT workforce shortage (projected at 220,000 unfilled roles) was already skewed toward AI and data roles. The APPI amendment adds regulatory tailwind to an already supply-constrained market. For engineers with ML, data engineering, or privacy engineering backgrounds, Japan's combination of high demand, lower competition relative to US and UK markets, and a now-permissive regulatory environment for AI development is genuinely unusual.

The April 2026 APPI amendment is significant enough to be worth understanding in full, including the parts that are less straightforward to welcome. Here is the complete picture as of late May 2026.

What the bill contains:

New consent exemptions for AI and research use. Personal data can be collected, processed, and shared with third parties without individual consent when used to produce statistical information. The government has interpreted "statistical information" broadly to include AI model training. The required safeguards are pseudonymization or equivalent anonymization, documented purpose limitations, contractual protections with data recipients, and public disclosure of data handling practices.

Specific Biometric Personal Information. A new protected category covering facial recognition data and fingerprint data converted to numerical format. This is the kind of data collected from foreign nationals at Japan's border entry. Companies handling this category must notify individuals or make information publicly accessible about how biometric data is being used, disclose what physical features are being converted and for what purpose, and allow individuals to request suspension of use or third-party transfer without the usual threshold requirements. Biometric data in this category cannot be shared with third parties via the standard opt-out mechanism.

Children's data. Parental consent required for children under 16. Enhanced deletion and suspension rights apply to children's data.

Administrative fines, for the first time. Japan's APPI previously had no administrative fine mechanism, with enforcement relying on corrective orders alone. The amendment introduces fines, with implementation expected by 2028.

Cross-border transfer reinforcement. Companies transferring personal data overseas for AI training on foreign infrastructure must conduct transfer risk assessments, implement contractual safeguards, and maintain audit trails. Japan is tightening oversight of how data leaves the country even as it loosens rules on domestic use.

The consent question:

It is worth being direct about what the consent exemption means in practice. Your publicly visible professional profile, employment history, and other personal data can be collected and used to train AI models without your knowledge or permission, provided the company applies the required safeguards. The pseudonymization requirement offers some protection, but re-identification from pseudonymized datasets is technically feasible, particularly with employment and demographic data. The "low risk" threshold that justifies consent-free use is assessed by the company doing the processing, not by a neutral authority or the individual whose data is being used.

The biometric data protections added in the same bill are meaningfully stronger, and their inclusion arguably highlights the gap in the general rule: the legislature clearly recognized that some data is sensitive enough to require explicit consent, while treating everything else as fair game for AI development by default.

You retain the right to request suspension of use or third-party transfer for your personal data. If you want to exercise that right, the company must publicly disclose how to do so. This is a weaker baseline than opt-in consent, but it is not nothing.

The government's stated goal:

Prime Minister Takaichi's administration has framed this as "Agile Governance": transparency and industry cooperation rather than prescriptive prohibition. The stated goal is to make Japan the best country in the world to develop and deploy AI. The APPI amendment is one pillar of that strategy alongside investment in compute infrastructure, AI research funding, and talent attraction policy.

Timeline:

Cabinet approval: April 7, 2026. Diet passage expected in the 2026 summer session. Full implementation within two years of promulgation, so by 2028 at the latest. Companies are adapting practices now; the full legal framework is not yet in place.

One thing that has not changed:

The Act on the Protection of Medical-Related Information (PIIFN) governs healthcare data separately and is not affected by the APPI amendment. Health and medical data remain under stricter consent requirements. The more permissive new rules apply primarily to employment data, behavioural data, financial data, and publicly available personal information.

Company: SmartHR, Inc.
HQ: Tokyo (Minato)
Category: HR Tech / Cloud SaaS / Workforce Management
English-Friendly: Yes, English is the primary engineering language
Hiring from Abroad: Yes, visa sponsorship, remote-first engineering

SmartHR is Japan's leading cloud-based HR software company. Their platform handles employment record management, social insurance filings, payroll, and workforce analytics for more than 60,000 organisations across Japan, from small businesses to major enterprise clients. They have been valued at over ¥500 billion (roughly $3.5B USD), making them one of Japan's most prominent SaaS unicorns.

The connection to this week's topic is exact: SmartHR processes employment records, social insurance data, and personnel information at scale, precisely the category of data that Japan's APPI amendment opens up for AI development. SmartHR already operates a suite of AI-powered HR features, and the new regulatory environment gives them significantly more room to build. If you want to work on AI products being built in Japan right now on a real regulatory frontier, SmartHR is a particularly good place to be looking.

Why they're relevant right now:

SmartHR sits at the intersection of three forces converging this year: Japan's APPI amendment (which benefits HR data companies directly), the mid-career hiring boom (which is creating demand for better HR and workforce management tools), and the broader AI product wave following the regulatory change. They are a funded, growing company building on a genuinely interesting dataset.

What stands out:

Current open areas include backend (Ruby, Go), frontend (TypeScript/React), mobile, data engineering, ML infrastructure, and platform/security engineering.

✏️ Application tip: SmartHR's interview process emphasises product thinking alongside technical depth. Given the HR data domain, they value engineers who think carefully about correctness, compliance, and data sensitivity rather than just engineers who can ship fast. If you have experience building systems where data accuracy and audit trails matter (financial, compliance, or medical-adjacent systems), lead with that. SmartHR's clients are using the platform for legally significant HR filings, so precision is a core value. The process is conducted in English.

🔗 Careers page

If your goal is to actually start applying, this is the approach I’d recommend.

Instead of relying on fixed job lists or jumping between job boards, you can search across company career pages directly and focus only on the roles that actually fit your situation. You can also set alerts, so new opportunities come to you instead of constantly checking.

You can start here

If you’re still figuring out your strategy or trying to understand how the market works, that’s what this newsletter is for.