MICE Explained: Four Motives Behind Espionage and How Employers Can Protect Staff

Hello, this is Ryuta Hamamoto from TIMEWELL.

You may have come across the word MICE in discussions of espionage and counterintelligence. It is an acronym for four motives that can lead a person to share information or cooperate with an outside party: Money, Ideology, Coercion or Compromise, and Ego. It has nothing to do with rodents.

Here is the conclusion up front. MICE is not a tool for spotting dangerous people. It is a map that helps an organization close, ahead of time, the gaps through which a person might say yes to an outside approach. If you use it to hunt for suspects, you will only spread suspicion through the workplace, and your defenses will not get stronger. In this article I trace where the term comes from, as far as I could verify, and then turn each of the four motives into a design that supports employees.

You will learn three things. How far back the term can be traced. What each of the four motives looks like in everyday terms. And how to build early signs and protections so that they help people rather than point a finger at them. The step-by-step procedures in the Japanese government's technology leakage guidance are covered in a separate article, so I leave them out here. If you would rather see where your company stands first, the Japanese edition of this article links to a three-minute self-check.

Seven terms to settle first

Discussions in this field often drift because people use the same words differently. Here are the terms used in this article, next to a picture from everyday life.

Term	Meaning	Everyday picture
Espionage	Efforts by a state or organization to collect another party's secrets through people or technology	Trying every trick to learn the recipe of the restaurant next door
Counterintelligence	Protecting your own information and people from espionage	Locking your recipe book and managing who comes and goes
Recruitment	Getting someone to provide information or cooperate	Asking, "Could you help me with something?"
Insider risk	The chance that information leaves through someone inside the organization	A person with a house key opening the door, unintentionally or under pressure
Trade secret	Non-public business information that is managed as secret (Japan's Unfair Competition Prevention Act)	A design kept in a locked safe
Early sign	A small change in behavior or circumstances. It proves nothing by itself	A change in the color of the sky, which promises no rain
Protective design	Reducing accidents through systems, not through individual vigilance	Removing the step in the floor rather than telling more people to watch their feet

The last two carry this whole article. Signs prove nothing, and protection comes from design, not from asking individuals to be careful. Keep those two points and MICE becomes a useful framework.

Where the term comes from: what I could and could not verify

I want to be clear first about how far the origin can be traced, because a shaky story that slips into a working document is hard to correct later.

Here is what I could confirm. A 2014 paper by the psychiatrist David Charney and John Irvin, written for the journal of the Association of Former Intelligence Officers, says the framework was probably popular as common sense before it was ever printed. It says the first appearance in print seems to be in a book by Stanislav Levchenko, a former KGB major who defected to the United States in 1979. In his memoir, On the Wrong Side: My Life in the KGB (Pergamon, 1988), he suggested four general motives for espionage: money, ideology, compromise or coercion, and ego.¹²

To be honest about what I read, I read that paper, which is a draft version, and not Levchenko's book. That is why I say "appears to." You will also see claims online that the CIA devised the framework. I could not support that with the sources I checked. I also confirmed that the FBI has a handbook on insider threats, but its site returned a 403 to automated access and I could not read it, so I say nothing about its content.³

What matters more is how later research treated the four categories. According to the same paper, the U.S. Department of Defense's Personnel Security Research Center (PERSEREC) published a report in 1992 based on a database of 120 espionage cases by Americans since World War II, and it identified six motivations. Beyond money, ideology and coercion, it added disgruntlement or revenge, ingratiation, and thrills or self-importance. The paper describes the last three as finer-grained versions of the Ego in MICE.² PERSEREC kept updating the data. Its 2017 report (Technical Report 17-10, by Katherine Herbig) compares 209 people across three periods starting in 1947. I could only confirm the bibliographic details, not the figures inside, so I do not quote numbers here.⁴

The same authors state the limits plainly. MICE is still widely cited, but it has limited value in predicting who will or will not act. The categories are too general to capture the complex motives of an individual, and using them risks pasting simple labels on people.² The insider-risk material from Britain's National Protective Security Authority (NPSA, formerly CPNI) is, as far as I could confirm from search excerpts, on the same side: motives are varied and often combined, and one should look at circumstances, not motive alone. I could not read the full text, so I limit myself to that excerpt.⁵

What about Japan's public material? I looked through the National Police Agency, the Shiga Prefectural Police and the Public Security Intelligence Agency, and found no material that uses the term MICE itself. The closest framing is written from the side of methods. The National Police Agency divides technology leakage into three patterns (cyberattacks, espionage, and leakage through economic or academic activity) and explains that, in the case of espionage, information can be taken through people and not only through cyberspace.⁶ I checked the original of the Ministry of Economy, Trade and Industry's Technology Leakage Prevention Guidance (Second Edition, 182 pages) and found no passage that sets out motives as a systematic typology like MICE. The case studies in Chapter 3 instead describe how contact progresses: dissatisfaction with pay, the pleasure of being noticed, and difficulty refusing after already accepting meals or gifts (Chapter 3, leakage cases, pp. 59-63).

Turning the four motives into everyday circumstances

Imagine a stranger on the street asking, "Could you carry this bag for me?" People who say yes do so for one of four reasons. They are paid for it. They agree with the person asking. They cannot refuse because the other person holds something over them. Or they are pleased to be asked. The four letters of MICE correspond to those four reasons.

Money is the easiest to understand, but it does not only mean greed. In the paper's account of a CIA research psychologist's view, the money motive ranges from strong materialism to the need to cover a family member's medical care or a child's education.² In many cases the doorway is not malice but a practical difficulty in life: debt, caring for a parent, a mortgage, a side business that failed. What an organization should look at is not the individual's wallet but whether a person who is struggling has somewhere to turn.

Ideology means acting on a belief about how the world should be. It is not only about states and politics. A researcher may believe that scientific results should be open to all. An engineer may believe a technology should be used for the good of the world. Beliefs like these can be the starting point, and the line between well-meant sharing and removing protected information can be invisible even to the person involved. An organization must never investigate or evaluate an employee's beliefs. But it does bear the responsibility of explaining what is protected and why.

Coercion or compromise is when a person acts not by choice but because someone holds something that would hurt them if exposed. The authors describe it as a form in which the person does not act of their own free will but out of fear of punishment or exposure. They add that it is the least reliable method of recruitment, because the person cooperates only as far as necessary and will try to break free when they can.² The person to protect here is the employee who could become a victim. If someone whose past mistake, family situation or private vulnerability has been used against them can take it to the company, this motive loses its hold.

Ego covers the wish to be recognized or to feel special. Levchenko meant a hunger for challenge, adventure and excitement. Later research broke it down into dissatisfaction with evaluation, a feeling of being underrated, and a wish to show off expertise.² What is distinctive here is that the person usually feels the grievance is legitimate. The unfairness may be real or it may be perceived. Either way, a transparent evaluation process and a way to get an explanation that makes sense are what help.

Side by side, the four look like this.

Motive	Everyday circumstance	How an approacher gets in	What an organization can do first
Money	Practical difficulties, worry about the future, an offer of attractive side income	Fees, well-paid side jobs, advisory roles	A confidential help desk for life and money matters, side-work and honorarium disclosure, rules on gifts and hospitality
Ideology	A sincere wish to do good, unease with the organization's direction	Approaches that appeal to shared values	Explain what is protected and why, document the line between open and restricted, a place to consult before publishing
Coercion or compromise	Past mistakes, family matters, things one would rather not have known	Demands backed by a hidden secret, step-by-step entanglement	A no-blame consultation channel, a written policy on self-reporting, a bridge to lawyers or the police when needed
Ego	Wanting recognition, frustration with evaluation, wanting to speak as an expert	Praise, requests, special treatment	Fair evaluation, recognition of contributions, a safe outlet for public speaking with a consultation rule

One more point. The four mix inside a single person. Someone short of money may also enjoy being relied on, and may then become unable to turn back because of a secret that was created along the way. The paper and the PERSEREC-based research both stress that motives combine and change over time.² Sorting people neatly into four boxes to diagnose them is outside what the framework was meant for.

How approaches actually unfold, and where each job is exposed

Next, what an approach looks like in practice, within what public material shows.

A page published by the foreign affairs division of the Shiga Prefectural Police (updated April 9, 2026) describes a former employee of a telecommunications company who, in February and March 2019, was incited by a person claiming to be a foreign government official and improperly obtained his employer's trade secrets. The contact moved in stages. The person first proposed providing information that was easy to hand over, then gradually asked for information related to trade secrets, and then mixed incitement, rewards and threats to steer the person along.⁷ The page also lists, as methods, approaches through social media and offers of rewards, instruction in concealment, threats and coercion, and joint research, funding, recruitment of talent and acceptance of students or visiting researchers.⁷

My own reading is that the early stages work on money and on ego, the pleasant feeling of being relied on, and the later stages work on vulnerability. Making the first request small is a way to leave no reason to refuse. That is exactly why consultation matters. If the person can raise it inside the company at the first contact, the later stages never happen. The National Police Agency likewise urges people to look closely at who they are dealing with, to stop and think, and to share and consult about even minor matters with a supervisor or colleague. It says that a message from a foreign company the person has no connection to on a personal social media account, or a present or meal offered as thanks, can be a sign of an approach.⁶

On online approaches, the British government in May 2022 released an app under the "Think Before You Link" initiative, from the Cabinet Office and CPNI, that helps people judge whether a profile is fake. The Director General of MI5 said the service had seen more than 10,000 approaches under disguised identities on professional networking sites, and the government urged caution toward unsolicited offers such as well-paid consultancy work.⁸ It is not a Japanese case, but the pattern is the same.

Leakage by people is only one route among several. For the whole map, including cyberattacks and equipment, see the article that breaks leakage into seven routes.

So is "this doesn't concern my company" a reasonable feeling? What follows is a general summary based on public material. Whether you become a target depends not on your size or name but on the information you hold and the roles of the people who touch it. The entrances differ by job.

Job	Possible entrances (general)	Examples of requests wearing a legitimate face
Researchers and engineers	Conferences, joint research proposals, requests to review papers or give talks, invitations as a visiting researcher	A paid lecture, a request for materials framed as academic exchange
Sales and business development	Trade shows, inquiries posing as customers, quotation requests	Detailed questions about specifications or manufacturing conditions
Procurement and purchasing	Contact through suppliers and agents	A quotation request that presupposes sharing drawings or specifications
Information systems	Proximity to administrator rights, relationships with maintenance and outsourcing vendors, high-paying side jobs posted as recruitment	Providing configuration details for a "simple task"
HR, general affairs and executive assistants	Job applications, visitor handling, handling of rosters and schedules	Questions about the org chart or travel plans

Each of these entrances looks like an ordinary request that blends into everyday work. That is exactly why a habit of sharing instead of deciding alone works.

Protective design by motive: treat signs as a reason to check in

This is the core of the article. For each motive I list the early signs an organization may notice and the protections it can put in place first. The column of signs is not a way to assess people. Think of it as a doorway for offering support.

Motive	Early signs (a doorway for the organization)	Protective design
Money	The person mentions a practical difficulty on their own. An unusual honorarium or gift is disclosed. Side-work disclosures and reality do not match	A help desk for life and money matters (an outside counseling service is fine). A disclosure system for side work and honoraria. Rules for accepting gifts and hospitality
Ideology	Unease about how widely research can be shared comes up as a work question. The person does not seem to accept why something is confidential	A patient explanation of what is protected and why. A written line between open and restricted. A contact point to consult before papers or talks
Coercion or compromise	The person reports "I had an uncomfortable contact." The person meets an outside party alone repeatedly with no explanation	A no-blame consultation channel. A written policy of not penalizing people who report themselves. A bridge to lawyers or the police when needed
Ego	Strong dissatisfaction about evaluation or treatment is expressed. The person speaks more often outside as an expert	Published evaluation criteria and explanations that make sense. Recognition of contributions. A safe place to speak publicly, with a rule to consult beforehand

Handle the Money row with special care. A company that investigates employees' household finances or debts is invading their privacy and will destroy trust. The row lists only what the person says voluntarily and what is disclosed through a system. The point is to provide an easy place to talk, not to dig. Getting that direction right matters most.

Some designs work whatever the motive. The first is least-privilege access. If you narrow who can touch what to the necessary range, then even under temptation or threat there is less that can be handed over. The second is the exit and transfer procedure: cut access on the last day, collect issued devices and media, and confirm the confidentiality obligation in writing. The third is a contact point for outside approaches. When someone outside offers work or a reward, people need to know whom to tell. For the concrete steps on permissions and removal routes, see our walkthrough of the technology leakage guidance, Second Edition. For the view of hiring through retirement as one line, see the article that connects the human phases. I will not repeat them here.

A word on internal reporting and the facts of the system. Article 11 of Japan's Whistleblower Protection Act requires a business operator to designate people to handle whistleblowing reports and to establish the systems needed to respond properly. For operators with 300 or fewer regular employees, this is an obligation to make efforts.⁹ The Act covers reports of facts subject to reporting, which means legal violations. A personal difficulty, or a consultation about a suspicious outside contact, may not fit that frame as it stands. I recommend choosing one more entry point, separate from the reporting channel, where people can bring anything. Call it a consultation desk if you like. What matters is that it can take in a concern before it has been shaped into a formal report.

You can tell whether this design works by whether it is used. A desk with no consultations at all is not evidence of safety. It may be a sign that the desk is hard to use.

How to treat signs: to protect, not to suspect

Here are the things not to do. This is the most important section of the article.

Do not turn the detection of a sign into grounds for suspecting a particular person. There are three reasons. First, most signs are ordinary troubles: a change in life circumstances, dissatisfaction with evaluation, fatigue. They happen to everyone, and nearly all have nothing to do with leakage. Second, the framework itself has limits. The researchers say that MICE has limited value in predicting who will act.² Third, a mistaken suspicion hurts the very employees you want to protect. A suspected person is isolated and stops asking for help when they truly need it. That is the situation most convenient for the party making the approach.

The practical principles are these. The first thing someone who notices a sign does is not to investigate but to speak. Ask, "How have you been?" and, if there is a difficulty, connect the person to the desk. Do not reflect it in evaluation or treatment, and if you keep a record, limit its purpose and who can read it. Do not impose a personnel penalty on the basis of a single sign. Do not use beliefs, nationality, origin, religion or family circumstances as risk indicators. The moment you do, it becomes discrimination, and it can be legally problematic.

Let me also speak about countries and regions. The four motives are human circumstances that can arise in any person in any country. Nothing in the public material I reviewed gives a basis for regarding people from a particular country or ethnic background, international students or employees with foreign nationality as a group more likely to be recruited. What to examine is not the attributes of a person but the route of an approach and the organization's own systems. There are situations, such as the "specific categories" in Japan's deemed-export rules, in which a check required by the system must be done. That is a procedure carried out according to the rules and within the scope related to the job, and it is separate from suspicion of an individual. For details see the article on deemed exports.

The same attitude is needed when citing real cases. In the case reported as the indictment of a former Asahi Kasei employee, Jiji Press and Kyodo News reported, citing the Aichi prefectural police, that the former employee said he disclosed information "to show that he was knowledgeable and experienced."¹⁰ But this is reporting at the indictment stage. A guilty verdict has not been reached, and I could not confirm whether the person admitted or denied the charge. It is not a finding of motive. The course of events is summarized in another article. The only general point I draw is that information can leave an organization without money being the motive. A single person's case must never be extended into a judgment about a company or a country.

On trips, on social media and in hiring interviews

Finally, three everyday settings.

On overseas trips and at conferences, someone may ask, "Could you send me the materials?" Do not agree on the spot. Take the request back, share it inside the company and then decide. If the organization sets this as a rule, the burden on the individual disappears, because the person can say, "Company policy is that I check first." The article on export control for business trips and online meetings is also useful.

On social media, confirm who a stranger is before responding. The National Police Agency asks people to stop and be careful when writing personal information in places that many people can see.⁶ If a profile on a professional network looks too polished, or a well-paid job offer comes without warning, do not reply alone. Share it. The kind of support released by the British government for checking whether a profile is genuine is a useful reference.⁸

In hiring interviews and when inviting outside experts, both sides need a checking mechanism. For applicants, verify their background fairly, with their consent and within the scope related to the job. The principle is not to screen by nationality or origin but to set a procedure for people who will handle important technology. The relationship between hiring checks and insider risk is covered from another angle in the article on the OpenAI recording-device case. You need both checks before people come in and an environment where they can consult once they are in.

Common misunderstandings, and a first step

Let me sum up in the form of answers to beliefs I often hear in practice.

"We're too small to matter" is dangerous. Whether you are a target depends not on size but on the technology and information you hold and the roles of those who touch it. The National Police Agency explains that companies and universities with leading-edge technology, regardless of size, can be targeted from abroad.⁶

"If we know the motive, we can spot the risky person" goes beyond what the framework was built for. As I wrote above, its own researchers say its predictive value is limited.

"A reporting channel is enough" is also a misunderstanding. The Act's reportable matters are legal violations, and a personal difficulty or a consultation about a suspicious contact often does not look like a report.

"It is only about people who need money" is wrong too. The doors of ideology, vulnerability and ego are open to people who have no financial trouble. Conversely, money trouble does not by itself lead to leakage.

A first step is to decide something small. This week, choose one place where people can raise difficulties or suspicious contacts without being blamed, and tell everyone. Next week, list the information you protect and narrow access to the people who need it. This month, test the procedure for cutting access at exit and transfer against a recent real case. The National Police Agency accepts consultations about technology leakage, and each prefectural police has a contact point.⁶

TRAFEED, our product, is an AI agent for export control classification and counterparty screening. It checks counterparties and joint-research partners and leaves the basis for each judgment on record. It does not assess the motives or inner state of employees. The final determination is made by your own export control manager. You can read about the service on the TRAFEED page, and consultations about building your program are available through a personal consultation.

One last personal thought. The value of learning MICE does not lie in sharpening an eye for suspicion. It lies in knowing what makes people waver and building an organization that does not leave a wavering person alone. Protective design is, at heart, a way of expressing trust in your employees.

References

Footnotes

1. Stanislav Levchenko, On the Wrong Side: My Life in the KGB (New York: Pergamon, 1988). The original was not read. This follows the account in the Charney and Irvin paper below. ↩

2. David L. Charney, M.D. and John A. Irvin, "A Guide to the Psychology of Espionage" (draft for the AFIO Intelligencer, version dated May 12, 2014). https://cyberwar.nl/d/20140500_CHARNEY_The_Psychology_of_Espionage_DRAFT_2014May12.pdf — A draft; the final version was not checked. References to the 1992 PERSEREC report and to Thompson (2009), Why Espionage Happens, follow this paper's account. English (checked 2026-11-06). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

3. FBI, "The Insider Threat: An Introduction to Detecting and Deterring an Insider Spy." Its existence was confirmed from the listing on the FBI counterintelligence page (archived copy: https://webarchive.library.unt.edu/web/20210119195659/https://www.fbi.gov/investigate/counterintelligence). The FBI site returned 403 to automated access and the text was not read (checked 2026-11-06). ↩

4. Katherine L. Herbig, The Expanding Spectrum of Espionage by Americans, 1947–2015 (PERSEREC Technical Report 17-10, August 2017). https://irp.fas.org/eprint/spectrum.pdf — Only bibliographic details (209 individuals, three periods) were confirmed from search results. The figures inside were not read (checked 2026-11-06). English. ↩

5. UK National Protective Security Authority (NPSA), insider risk material. https://www.npsa.gov.uk/blog/if-you-have-people-you-have-insider-risk-david-smith-case-study-0 and others — The site returned 403 to automated access and the full text was not read. This is limited to what could be confirmed from search excerpts (motives are mixed, and context matters as well as motive) (checked 2026-11-06). English. ↩

6. National Police Agency of Japan, "Economic Security" (prevention of technology leakage, leakage through espionage, countermeasures, consultation contact keizaianpo110@npa.go.jp, telephone +81-3-3581-0141). https://www.npa.go.jp/bureau/security/economic-security/ — Videos and pamphlets were published in January 2026, with earlier ones in November 2023 and September 2022. In Japanese (checked 2026-11-06). ↩ ↩² ↩³ ↩⁴ ↩⁵

7. Shiga Prefectural Police Headquarters, Security Department, Foreign Affairs Division, "Espionage and Leakage of Technical Information" (updated April 9, 2026). https://www.pref.shiga.lg.jp/police/pa16/15932.html — The case concerns an approach to a former employee of a telecommunications company by a person who claimed to be a foreign government official. The country is not named in the source. In Japanese (checked 2026-11-06). ↩ ↩²

8. UK Government (Cabinet Office and CPNI), "New app to counter malicious approaches online" (May 17, 2022). https://www.gov.uk/government/news/new-app-to-counter-malicious-approaches-online — Includes the remark by the MI5 Director General (more than 10,000 disguised approaches on professional networking sites). English (checked 2026-11-06). ↩ ↩²

9. Whistleblower Protection Act (Act No. 122 of 2004), Article 11. https://laws.e-gov.go.jp/law/416AC0000000122 — The text was checked through the e-Gov Law Search API. In Japanese (checked 2026-11-06). ↩

10. Jiji Press, "Disclosed information 'to show he was knowledgeable': former Asahi Kasei employee indicted, Nagoya District Public Prosecutors Office" (October 27, 2026). https://www.jiji.com/jc/article?k=2026102701077&g=soc ; Kyodo News, "Former Asahi Kasei employee indicted over leak of trade secrets" (same date, Yahoo! News Japan). https://news.yahoo.co.jp/articles/6a329a93dc1df0878cdb0ab226617c1ca2ac3ba0 — Reports at the indictment stage; the statement was reported as coming from the Aichi prefectural police. No verdict has been reached and whether the person admitted or denied the charge was not confirmed. In Japanese (checked 2026-11-06). ↩