Labour market shifts and the evolution of trade secret protection under Japan’s Unfair Competition Prevention Act

Takafumi Ochiai
Atsumi & Sakai, Tokyo
takafumi.ochiai@aplaw.jp

Yuka Daimon
Atsumi & Sakai, Tokyo
yuka.daimon@aplaw.jp

A changing workforce and a dual policy response

Japan’s labour market is at a turning point. While elements of the traditional lifetime employment model persist, mid-career hiring has increased significantly, particularly in the technology sector and among skilled professionals. The government has continued to promote workforce mobility. In 2018, the Ministry of Health, Labour and Welfare.^[1] revised its recommended Model Employment Rules to establish a default principle permitting side jobs and concurrent employment. At the same time, the spread of remote work and cloud-based business tools, driven by technological change and accelerated by the Covid-19 pandemic, has fundamentally altered how and where employees access confidential information.

These structural changes have produced a measurable increase in trade secret disputes. Consultations at the National Center for Industrial Property Information and Training (INPIT) Trade Secret Advisory Office have risen steadily, and the National Police Agency reported 22 criminal trade secret cases involving 45 suspects in 2024.^[2] Civil litigation has also intensified, spanning not only traditional manufacturing, but also diverse sectors including telecommunications, food services and advanced materials. In Japan, the secrecy management requirement has historically been applied strictly, and trade secret claims have frequently been rejected. More recently, however, both case law and the Ministry of Trade, Economy and Industry (METI) Management Guidelines for Trade Secrets,^[3] which courts may refer to, have been interpreted more flexibly, with the authorities involved declining to deny trade secret status on the basis of minor administrative deficiencies where the employer’s overall intent to maintain secrecy is evident.

Japan’s policy response reflects a recognition of the need to promote workforce flexibility on one hand, while materially strengthening the legal infrastructure for trade secret protection on the other. The Unfair Competition Prevention Act (UCPA)^[4] was substantially amended in 2023, and the METI Management Guidelines for Trade Secrets were comprehensively revised in March 2025, the first major update in six years.

Trade secret protection is discussed not only in the context of the UCPA, but also within the framework of economic security. The Foreign Exchange and Foreign Trade Act (FEFTA)^[5] regulates the transfer of technologies with national security implications. Following a 2022 clarification of ‘deemed export’ controls, technology provided to foreign-national employees or researchers residing in Japan who are under the strong influence of a foreign government or foreign entity may require prior approval, not only in regard to cross-border transfers, but also domestic provision. In December 2024, the METI introduced an enhanced government-industry dialogue scheme under the FEFTA, establishing a mandatory prior reporting system for the overseas transfer of critical technologies.15 technologies are currently designated. Where concerns are not resolved through dialogue, the METI may issue an ‘inform’ requiring a formal licence application.^[6] The Act on the Promotion of Ensuring National Security Through Integrated Implementation of Economic Measures (2022)^[7] introduced a patent application non-disclosure system. The Act on the Protection and Utilisation of Critical Economic Security Information established a security clearance system for personnel handling critical economic security information, including information relating to cyber threats and supply chain vulnerabilities.^[8]

Among the cases discussed below, the AIST case — involving the misappropriation of sensitive research data by a foreign-national researcher at a government research institute — sits squarely at the intersection of trade secret and economic security concerns.

The 2023 UCPA amendment and the 2025 revision of the Management Guidelines for Trade Secrets

The UCPA defines a ‘trade secret’ as information that satisfies three requirements: (1) secrecy management (himitsu kanrisei), (2) usefulness and (3) a non-public nature. Of these, secrecy management is the most frequently contested element in litigation. The UCPA provides civil remedies (damages, injunctions) and criminal sanctions only for information meeting all three requirements. Broader confidential information is protected only through contractual mechanisms, such as non-disclosure agreements (NDAs) or general tort law.

The 2023 amendment (effective April 2024) addressed three dimensions.^[9] First, the scope of ‘limited-provision data’ protection was expanded through revised definitions and a new presumption of use, which originally targeted Big Data sharing services, but protection was extended to commercially valuable datasets managed as trade secrets. Second, Japanese courts were given jurisdiction over trade secret misappropriation committed abroad, which was a direct response to cases where departing employees take secrets to foreign competitors. Third, in regard to damages litigation, rightsholders may now claim royalty-equivalent amounts even for losses exceeding their own production capacity.

The March 2025 revision to the Management Guidelines responded to the practical environment in which these statutory provisions operate. The METI maintains two principal guidance documents: the Management Guidelines for Trade Secrets, which set the ‘minimum level of measures’ required for trade secret protection, and the Confidential Information Protection Handbook,^[10] which contains more advanced recommendations beyond the legal minimum. The revision of the Guidelines was informed by recent case law, particularly cases involving former employees, and by technological change. In a landscape where secrecy management is frequently at issue in departing-employee disputes, the revision represents a substantive strengthening of the framework for trade secret protection.

Relaxation of the secrecy management requirement: strengthening protection

Whether satisfaction of the secrecy management requirement demands proof of individual employee awareness and ‘reasonable classification’ (goriteki kubun) of trade secrets from other information has been a recurring issue in litigation. The revised Guidelines clarified that where secrecy management can be assessed objectively, including in reference to the awareness of the employee population as a whole, it is not necessary to prove recognition by each individual employee. Recognising that cloud service environments may limit an employer’s ability to control classification, the revision further clarified that reasonable classification is merely one evaluative factor in assessing secrecy management, rather than an independent requirement.

The revision also underscored the importance of establishing work rules covering employees and officers, conducting employee training and awareness programmes and concluding NDAs with business counterparties. It was clarified that in certain circumstances, technical access controls (such as ID and password protections) combined with normative measures (such as work rules and confidentiality pledges prohibiting disclosure) may suffice.

Taken together, the concept of ‘reasonable classification’ has been reclassified from a requirement to one of several evaluative factors. This, combined with the clarification that minor discrepancies from the detailed management measures set out in the Guidelines do not in themselves negate secrecy management, reduces the risk that technical or procedural deficiencies will defeat an otherwise genuine secrecy management regime.

Cybersecurity, generative artificial intelligence (AI) and technology-related considerations

The revised Guidelines also address the intersection of secrecy management with cybersecurity. According to the Information Technology Promotion Agency’s (IPA) 2024 survey, information leakages attributable to external cyber-attacks have increased significantly and have become a major factor alongside information leakages caused by departing employees.^[11] The Guidelines clarified, for instance, that even where trade secrets are leaked onto the dark web through a cyber-attack, the non-public status of such information is not automatically negated.

Similarly, where a trade secret consists of a combination of publicly available information, it may remain protected if the combination is not easily ascertainable or requires significant cost to compile. Trade secrets embodied in commercially sold products may also retain protected status where extraction would require specialised technology and a significant period of time.

The revised Guidelines also clarified a ‘management unit’ framework for generative AI. If trade secret information within management unit A is used in a generative AI system and the output appears under management unit B, secrecy management is not automatically negated, provided that each unit is subject to secrecy management and the information remains within the employer’s overall controlled environment.

In addition, the Confidential Information Protection Handbook warns against the risk of information leakage arising from inadequate organisational rules governing generative AI use, cautioning employers not to permit the use of AI services that lack appropriate contractual confidentiality protections. The Handbook also recommends the use of risk assessment and countermeasures addressing supply chain (subcontractor) risks, which is a frequent source of information security incidents at Japanese companies, as well as malicious AI-enabled threats, such as fake content, impersonation and phishing.

These regulatory developments are illustrated by recent case law. In the National Institute of Advanced Industrial Science and Technology (AIST) case, a foreign-national former researcher at the AIST was convicted of misappropriating fluorine compound synthesis know-how.^[12] The court held that AIST’s institutional-level measures, namely internal regulations governing research outputs, system access controls and employee training, collectively satisfied the secrecy management requirement, even in the absence of document-level ‘confidential’ tagging. This ‘comprehensive secrecy management’ approach reduces the evidentiary burden for organisations managing large volumes of technical information. In the SoftBank v Rakuten Mobile case, a sales engineer transferred 5G base station documents to their personal devices before joining rival firm Rakuten Mobile. A criminal conviction, including a two-year suspended prison sentence was finalised in July 2025.^[13] The case became widely known as a souvenir transfer (temiyage tenshoku), the practice of taking proprietary information as a parting gift when changing jobs. In the Kappa Create case, the former president of the Kappa Sushi restaurant chain was convicted of misappropriating a competitor’s product cost data, illustrating that trade secret risks extend well beyond the technology sector. Across industries, trade secret case law increasingly intersects with technology and digital governance issues.^[14]

Non-compete clauses: a calibrated complement

Beyond trade secret protection under the UCPA, post-employment non-compete clauses serve a complementary role. Japanese courts apply a six-factor proportionality test, derived from the 1970 Foseco Japan Limited case. The METI Confidential Information Protection Handbook, in its Reference Material 5,^[15] positions non-compete obligations as a tool to ‘support’ confidentiality obligations, which is applicable not only where the employee might disclose secrets, but also where the employee might use them. The protectable interest need not be limited to statutory ‘trade secrets’ under the UCPA: proprietary know-how that is difficult to manage as a formal trade secret may still justify the use of a non-compete clause where the employer can demonstrate a legitimate interest in its protection. Clauses of one year or less, limited to specific job functions, and supported by compensatory measures are more likely to be upheld. Those applied to employees without genuine access to protectable information, or imposing broad geographic restrictions, are more likely to be struck down.

In practice, employers increasingly deploy a dual-protection approach: a confidentiality agreement covering trade secrets and sensitive information, supplemented by a narrowly tailored, short-duration non-compete clause for employees with genuine access to critical know-how. As the labour market becomes more fluid and side jobs become more prevalent, this calibrated approach provides a middle road between jurisdictions that have moved towards outright bans on non-compete clauses and those that enforce them broadly.

Conclusion

Japan is pursuing workforce mobility and trade secret protection simultaneously. Not as contradictory goals, but as complementary dimensions of a competitive economy. The 2023 UCPA amendment and the 2025 revision to the Management Guidelines reflect a deliberate effort to reduce the barriers to employee movement and discourage the use of overly restrictive non-compete clause, while at the same time strengthening the legal and practical infrastructure for protecting trade secrets against the risks posed by the digital environment, including cloud computing, generative AI and cyber-attacks. For international practitioners advising on departing-employee matters involving Japan, this dual-track approach merits close attention.