Reverse Engineering Is Hard. Hiring the Engineer Is Not.
If the allegations hold, the Asahi KASEI cases show how Japan’s lead in semiconductor packing materials depends on people it struggles to protect.
On October seventh, a former employee of Asahi KASEI, a Japanese chemical manufacturing company, was arrested on suspicion of leaking semiconductor material information to a foreign company. Less than a month ago, this person was also arrested on suspicion of industrial espionage involving Shandong Shengquan New Materials, a Chinese chemical company. In the previous case, according to Asahi KASEI’s official statement, the former employee allegedly took confidential information related to its latent curing agent and sent it to Shandong Shengquan New Materials. In the latest case, the former employee allegedly took confidential information on photosensitive dielectric material in 2018, at the time of his departure, without returning the company’s hard disk drive, and sent it to a foreign company in 2024, after he left Asahi KASEI. If the allegations hold, the two arrests expose a weakness larger than one company’s lapse: in chipmaking materials, critical know-how is held by individual engineers, and Japan has few safeguards once they leave the company.
What’s the Material for?
The material in the October case, photosensitive dielectric material, is presumably Photosensitive Polyimide (PSPI), a kind of resin used in semiconductor (or chip) packaging, a process to assemble chips into an actual device and protect chips and wiring from outside dust and moisture. That’s not just a cover for the chip. Because all chips emit heat, it must meet strict specs such as thermostability and the coefficient of thermal expansion (CTE). The photosensitivity is trickier; the material needs to react to a specific kind of light and change its chemical structure to form the circuit pattern on it, which makes the technical barrier to entering this industry extremely high. Asahi KASEI has the top share in the PSPI market accounting for more than 50% of the market. The PSPI market is dominated by Japanese companies and four of the top five companies are Japanese and the fifth is a Japanese-American joint venture. These five companies hold more than 96% of the market share. Even if a chip company can make a top-notch chip such as a 2~3 nm chip, without PSPI, the chip is too exposed to the outside environment to be useful. Japanese companies set a de facto standard for chip packaging.
The Difference Between the Two Industries
Unlike the so-called chip industry or device industry, the chipmaking materials industry has more analog aspects from design to manufacturing. In contrast to the device field, where reverse-engineering is relatively easy because most of the circuit structure and materials are physically visible, chipmaking materials such as polymers leave fewer clues for reverse engineers. Polymers are long chains built from many repeating units (the name comes from the Greek for “many parts”). Working out how many units a given chain contains, and how they are arranged, is difficult even for a polymer engineer. Identifying the raw materials used to synthesise the polymer is harder still, because most of them are removed during purification. But these facts are not the only factor in reverse-engineering. In a chipmaking factory, a silicon wafer usually has to go through more than two thousand steps, and it takes a few months. On the other hand, polymer synthesis usually takes less than a week and involves fewer than a hundred steps. In other words, while an engineer in a chip factory knows only a tiny percentage of the big picture, an employee in a polymer factory has a much broader view of the product. These two aspects of the industries together indicate that, since it is naturally difficult to reverse engineer, the human can be a weak point.
Inferring China’s Intentions
The October case may yet fall apart. The allegation may be unfounded, the employee may have acted for money alone, and the foreign company may not be Chinese. Even so, Chinese interest in chipmaking materials would fit Beijing’s self-sufficiency drive. The 50% rule covers equipment, not materials, but it shows the direction. At the end of 2025, Reuters reported that Beijing imposed a rule requiring Chinese semiconductor companies to procure at least 50% of chipmaking equipment from domestic companies. While some Chinese semiconductor companies such as SMIC have successfully manufactured chips which are only a few generations behind the Western-made chips, the chipmaking equipment and materials are dominated by American and Japanese companies. Because of the difference in reverse engineering and because it makes more sense to access the information from human sources in the polymer industry than in the semiconductor industry, and given that the PSPI market is dominated by Japanese companies, China likely wants to break into the market. From a Chinese perspective, as China has used its rare earth export controls as an economic threat to Tokyo, Tokyo could use its dominant position in some industries in the same way.
Japan has been here before, and the most instructive precedent is in materials rather than chips. Grain-oriented electrical steel, a niche product used in transformers, took more than 20 years to develop. In 2012, Nippon Steel sued POSCO of South Korea for about 100 billion yen, alleging that POSCO had obtained its trade secrets through a former Nippon Steel engineer, who allegedly received several hundred million yen. The case surfaced through a separate trial in South Korea, where a former POSCO employee was convicted of selling electrical steel technology to China’s Baosteel and testified that POSCO itself had taken it from Nippon Steel. In 2015, POSCO paid 30 billion yen to settle. The know-how traveled in a person, not in a reverse-engineered sample.
What Happens to Retired Engineers?
Although many tech companies ask former employees not to use any information in the next job or even not to take a position at any competitor, there remains the freedom to choose one’s job in the market. While some companies can pay twice the industry average salary others cannot. An individual has the right to choose the employer. In an era in which information sometimes matters more than military power does, policymakers must be aware of these problems. Retirement ends a contract, not what an engineer knows.
My three proposals are as follows: First, treat retirement as a security event. Because a company hard disk drive stayed in a former manager’s hands for six years, firms should collect and audit devices before the last day and renew confidentiality undertakings. Second, Japan should conduct more thorough screening of critical materials.
The Act on the Promotion of Ensuring National Security Through Integrated Implementation of Economic Measures (2022) was updated in June 2026 and it designated semiconductors as critical materials, but niche products may fall outside its scope. Third, based on the screening results, Japan should provide financial incentives to those who have access to confidential technologies. Patriotism does not pay the monthly bills. The former employee is in his sixties and has changed jobs several times in the past six years.
The October case may still prove unfounded, and no foreign company may in fact hold PSI information. But economic security policy has to plan for the worst case, not the best. Japan’s chipmaking materials industry is deeply integrated into the chipmaking industries of partners such as Taiwan, South Korea and the United States, so industry espionage is not only Japan’s problem. Japan has spent decades perfecting materials that rivals cannot easily copy. It has spent far less thinking about the people who carry them in their heads.
