Which countries run the most information operations?

— Russia, the United States, China, and what Japan should be careful about

Nikkei published an article titled: “House of Representatives election: About 400 China-linked accounts conducted an ‘anti-Takaichi operation,’ using Japanese-language posts and AI in sophisticated ways” .

I felt a strong sense of discomfort reading it.

The reason is simple: the story strongly implies “information operation,” yet it does not sufficiently compare whether it actually meets that threshold—or where it sits in the international context.

Foreign information operations do exist. Russian interference in elections has been documented. Conspiracy ecosystems originating in the U.S. have crossed borders and spread internationally. China, too, engages in narrative-shaping efforts aimed at foreign audiences.

The real question is not “which country is the villain,” but what scale and what structure of influence we are looking at.

What does “400 accounts” mean in the context of X as a whole? How does it compare with Russia-related cases? And where does it sit when compared with the global spread of English-language conspiracy ecosystems?

And one more thing.

When we label something an “information operation,” are we truly judging it by behavior— or by who is being criticized?

I want to examine this question not through emotion, but by organizing it through international statistics and structural comparison.

1. Why is “China operations” the only focus?

Nikkei’s article infers “China-linked accounts” based on signals such as similar posting patterns, clusters of account-creation dates, a sudden shift into Japanese-language posting, and traces of Chinese-specific wording.

As data analysis, that has a certain level of plausibility.

But there is a question we should not skip.

Why was this framed so prominently as an “anti-Takaichi operation”?

According to the article, the targets were roughly 400 accounts. The activity concentrated during the Lower House election period, and it allegedly amplified narratives linking Ms. Takaichi to the former Unification Church.

Yet the issue of scale remains. Compared with the total number of Japanese-language active accounts on X, “400” is a very limited number. The article itself also notes that the overall impact on the election was “limited.”

And still, the headline foregrounds a powerful frame: “China-linked 400 accounts” and “anti-Takaichi operation.”

This raises another problem.

What exactly do we mean by “information operation”? Is it coordinated inauthentic behavior? Or is it political manipulation directly conducted by a state?

Nikkei’s body text leans toward the former. But the headline strongly pulls the reader toward the latter.

If this is about coordinated inauthentic behavior, then the proper focus should be the behavior, not the country label.

Once a headline emphasizes “China” as an attribute, the discussion tends to shift from behavioral analysis to country-based labeling.

The central question is not whether the actor is “China.”

It is whether our criteria for calling something an “operation” are consistent.

If the standard changes depending on who is being criticized, that is not information analysis—it is narrative formation.

2. Russia: State-led Polarization-Oriented Information Warfare

A defining characteristic of the Russian model lies in the strategic expansion of polarization as a matter of state policy.

Rather than directly securing victory for a specific candidate, the objective is often to amplify distrust and social confrontation, and to erode confidence in institutions.

Investigations surrounding the 2016 U.S. presidential election confirmed that Russia’s Internet Research Agency (IRA) operated large-scale inauthentic accounts, as documented in the Mueller Report and reports by the U.S. Senate Select Committee on Intelligence.

In congressional testimony in 2017, Facebook stated that IRA-related content may have reached approximately 126 million U.S. users.

Twitter announced that it had identified about 3,800 IRA-linked accounts, and that their posts exceeded 10 million tweets in total.

This was not a matter of a few hundred accounts. It was an information campaign conducted at the scale of thousands of accounts and tens of millions to over one hundred million in potential reach.

The method was not simply “pro-Trump.”

• Accounts inflaming racial tensions
• Pages emphasizing law-and-order themes
• Posts advocating anti-immigration positions
• Posts simultaneously amplifying pro-immigration messages

By stimulating both sides of ideological and social divisions, a structure aimed at placing society into a polarized state has been identified.

Since 2016, conspiracy narratives such as QAnon have also expanded rapidly.

Narratives such as the “Deep State” and “election fraud” were linked to the environment that culminated in the January 6, 2021 U.S. Capitol attack.

QAnon itself largely originated within the United States, but multiple studies have pointed out that Russia-linked accounts also amplified such content.

Whether Russian interference directly determined the election outcome has not been officially concluded.

However, many analyses argue that the expansion of polarization and institutional distrust ultimately created an information environment that was advantageous to then-candidate Trump.

The essential point is that the Russian model is characterized less by “making someone win” and more by a strategy of “breaking” societal cohesion.

When viewed against this scale and level of impact, the question naturally arises as to how activities involving only several hundred accounts should be positioned by comparison.

3. The United States: Less a State Actor, More a “Self-Amplifying Information Structure”

The structure of the United States differs from that of Russia or China.

What has become a major concern in recent years is not so much a state acting as a single unit to target foreign audiences, but rather the phenomenon in which domestically generated radicalization and conspiracy narratives self-amplify through platforms and then spread internationally.

A symbolic example is QAnon.

Narratives such as the “Deep State,” “massive election fraud,” and “elite conspiracy” spread rapidly through algorithms, and are widely analyzed as having influenced the environment surrounding the January 6, 2021 U.S. Capitol attack.

The defining feature of this phenomenon lies less in state direction and more in a structure where markets, partisan politics, and platform design converge.

• Algorithm-optimized “anger” content
• Sensationalist media driven by clicks and advertising revenue
• Monetization of extreme narratives via donations and subscription models
• Conspiracy narratives that reinforce partisan polarization
• The expansion of exclusionary rhetoric and scapegoating

Anger and fear are particularly easy to amplify.

Studies suggest that posts designed to strongly stimulate emotion tend to spread faster than neutral content, and are more likely to generate retweets and shares.

The Cambridge Analytica scandal in 2016 showed how data analytics and targeted advertising could be used to deliver political messaging with high precision.

In the 2020 U.S. presidential election, the “election fraud” narrative was largely amplified by domestic political actors, and then spread internationally.

The crucial point here is that, unlike the Russian model of “disruption from the outside,” the U.S. case has a domestic amplification mechanism that can become a source of influence itself.

In other words, even without external intervention, an environment can form in which algorithms and commercial incentives allow polarization to self-amplify.

This structure crosses borders easily.

English-language conspiracy narratives and inflammatory rhetoric are translated and flow into Japanese-language spaces as well.

That is not necessarily a state-led information operation. However, the result can still be influence on political discourse in other countries.

In this sense, the United States is closer not to a single “state actor,” but to a self-amplifying information structure.

4. China: Narrative Formation Model

China’s information activities differ in character from Russia’s strategy of maximizing polarization.

A more distinctive feature is the concentration on specific themes, the formation of a “desired story” (narrative), and the effort to establish that narrative over the long term.

However, it is important to note that in China’s case, even when described as “state-led,” this does not necessarily mean a single, unified chain of command operating in a monolithic manner.

International research and governmental reports often describe the Chinese model as layered, involving state actors, quasi-state actors, and private actors.

• Official model: Statements from embassies, foreign ministry accounts, and state media
• Quasi-official model: Amplification of narratives through government-affiliated groups and influencers
• Outsourced model: Spread via PR firms and digital marketing companies (with lower transparency)
• Mobilized participation model: So-called “50-cent party” style mobilization, where individuals create an atmosphere (mixing authenticity and coordination)

In some cases, it may be possible to conclude that “the state directly ordered it.” In others, actors aligned with state positions amplify messages in a looser, more decentralized manner.

Unlike the Russian model, which often inflames both sides simultaneously to destabilize society, the Chinese model more often moves in one direction, seeking to reinforce a particular frame or position.

Themes tend to be relatively concentrated.

• Improving China’s international image (economic cooperation, peaceful development, etc.)
• Claims of legitimacy regarding Taiwan
• Counter-narratives to criticism over Hong Kong and Xinjiang
• Justification of positions concerning the South China Sea

The technique used is often not short-term explosive spread, but repetition of the same themes to stabilize recognition over time.

In international comparative contexts, the Chinese model can therefore be characterized less as “viral explosion” and more as reinforcement of specific perceptions and framing, as well as counter-criticism.

In recent years, multilingual dissemination has increased.

Messages are communicated not only in English, but also in Japanese, Spanish, and other languages, with translation used to deploy consistent messaging.

The recent article referenced earlier infers “China-linked accounts” based on elements such as sudden increases in Japanese-language posts and expressions characteristic of Chinese usage.

However, an important distinction must be made: the presence of “traces of Chinese language usage” is not equivalent to proof that “the state directed the activity.”

Chinese-language personal accounts, overseas Chinese communities, politically aligned online groups, outsourced operators, and official state accounts— Chinese-origin messaging operates across multiple layers, and its meaning changes depending on which layer is involved.

Therefore, the focus of analysis should not be simply whether something is “Chinese,” but rather what kind of behavior occurred, to what degree it was coordinated or organized, and with what objective.

To summarize:

• Russian model: Polarization maximization (disruptive model)
• U.S. model: Internal ecosystem amplification (market + polarization + platform design)
• Chinese model: Narrative formation with overlapping state and quasi-state layers (constructive model)

Given these three types, the question becomes: where should activities involving only several hundred accounts be positioned?

In the next chapter, we compare scale and impact, and place the “400 accounts” within an international analytical context.

5. Where Do the “400 Accounts” Stand Internationally?

So far, we have examined three models: Russia, the United States, and China.

Where, then, do the approximately 400 accounts identified in the recent article stand within the broader international context of information operations?

In the 2016 U.S. presidential election, approximately 3,800 accounts were identified as linked to Russia (according to Twitter’s official disclosure).

Facebook testified that related content may have reached up to 126 million users.

This was an operation conducted at the scale of thousands of accounts and tens of millions to over one hundred million in potential reach.

Similarly, campaigns tracked by the EU’s External Action Service (EEAS) have confirmed sustained, multilingual, cross-platform dissemination in Russian-linked operations.

Compared with such cases, a figure of 400 accounts is limited in scale.

Even the article itself notes that the overall electoral impact was “limited.”

Of course, a smaller scale does not mean it should be ignored.

However, treating it as equivalent to major international interference cases would overlook a clear difference in magnitude.

More importantly, the number of accounts and their impact do not necessarily correlate proportionally.

In the Russian model, organized coordination, advertising spending, and algorithmic optimization significantly amplified reach.

In the case of the 400 accounts discussed here, the article does not present evidence of such advertising operations or large-scale network coordination.

Even if influence is acknowledged, it is necessary to carefully distinguish whether this constitutes large-scale state-level strategic intervention or more limited-scale activity.

In international comparison, the landscape can be broadly grouped into:

• Persistent operations run at the scale of thousands to tens of thousands of accounts (the “Russia-type” model)
• An “ecosystem-type” model in which domestic structures function as amplification mechanisms (the “U.S.-type” model)
• A theme-focused model in which state-linked actors may be mixed in (the “China-type” model)

Against these, a cluster of around 400 accounts is likely to differ in both scale and character from large-scale, high-impact operations.

The key question is not scale alone, but how we position it—and how we talk about it.

If a small case is overstated as a national-level threat, public debate loses its composure. Conversely, if it is dismissed simply because it is small, we loosen the standards of verification.

What is needed is neither fear-driven rhetoric nor denial, but evaluation grounded in comparison.

This comparison does not deny the existence of foreign influence. Rather, it is an attempt to place its scale and structure within an international context.

6. Domestic Structures Within Japan

Focusing only on external actors risks overlooking essential elements.

Polarization and radicalization are not necessarily imported from abroad.

Within Japan as well, structures exist that amplify information.

For example:

• Sudden surges of specific keywords during election periods
• Mutual amplification between anonymous aggregation sites and social media
• Simultaneous mass posting of identical arguments
• High-volume production of extreme narratives (exclusionism, conspiracy theories, hate speech)

These phenomena are not necessarily foreign information operations.

However, when combined with algorithmic structures and advertising models, they can rapidly become visible and widely disseminated.

Furthermore, the expansion of government and ruling-party publicity and advertising budgets is itself a structural factor that influences the information space.

In recent years, digital advertising and public relations spending have increased, creating an environment in which policy messages and political claims are distributed at scale across platforms.

This is neither illegal nor inherently improper.

However, a structure in which actors with greater financial resources can expand their exposure inevitably affects the balance of public discourse.

When discussing several hundred foreign-linked accounts, if we fail to simultaneously consider domestic funding, advertising capacity, and amplification structures, the full picture cannot be seen.

Particular caution is needed regarding narratives that claim, “someone is manipulating everything.”

Such narratives can function as devices that simplify complex analysis.

By reducing multifaceted social phenomena to a single external force, debates become easier to understand — but domestic structural factors recede from scrutiny.

The information space is formed through the interaction of:

• External narratives
• Domestic political polarization
• Commercial incentives
• Algorithmic optimization
• Disparities in financial power and exposure

Even if there is an external trigger, whether it spreads and becomes established depends on domestic structures.

Impact is determined not only by “the volume from outside.”

It also depends on a society’s history of polarization, trust in institutions, media literacy, and the quality of political discourse.

Overemphasizing external threats while neglecting domestic structural issues creates greater long-term risk.

What is required is not an emotional response about the presence or absence of foreign interference, but a calm discussion about how amplification structures themselves should be redesigned.

7. What Japan Should Be Careful About When Publishing This Kind of Coverage

Reporting on information operations is essential for society. Transparency and verification are among democracy’s core safeguards.

At the same time, the way such stories are framed can itself be used to form new narratives.

Communicating that external threats exist is not the same as simplifying external threats into a single storyline. When Japan discusses this topic, there are three points that require particular caution.

① The trap of single-country “demonization”

Simplifications such as “Everything is China” or “Everything is Russia” are easy to grasp—but dangerous.

Real-world information spaces are shaped by multiple overlapping factors: state-directed activity, commercial incentives, market- and algorithm-driven amplification, and domestic political polarization.

Reducing these dynamics to a single country obscures complex causal relationships and reproduces a different kind of “story.”

A frame that depicts one country as the clear “enemy” may feel intuitive in the short term, but over the long term it can lower analytic precision and turn political judgment into emotional reflex.

② The danger of turning criticism into “foreign influence”

The tendency to label domestic criticism or dissent as “foreign information operations” shrinks the space for free discussion.

In a democracy, criticism functions as a self-correcting mechanism.

If criticism of the government—or opposition to policy—can be treated as “the enemy” simply because it is said to be “foreign” or “externally influenced,” healthy debate becomes harder to sustain.

The mere fact that a message originated abroad is not proof of an operation.

What matters is whether specific behaviors are present: inauthenticity, coordination, deception, and systematic amplification.

If the standard changes depending on political “stance,” that is not verification—it is selection.

③ Fear narratives can be used to justify domestic control

Emphasizing external threats can become a tool to legitimize stronger domestic control.

• Expansion of restrictions on expression
• Strengthening of surveillance powers
• Administrative or informal intervention into dissent

National security is, of course, important. But “the external enemy” narrative is politically powerful: fear accelerates narrative formation, while also encouraging society to skip careful scrutiny.

Historically, “external threat” narratives have functioned both to unify societies and to reduce pluralism.

That is why verifying threats and protecting freedom must be pursued together.

Even if there was an external trigger, we still need to measure calmly: what scale it reached, and what degree of impact it had.

At the same time, we must also examine how domestic structures and media framing amplify that narrative.

Rather than repeatedly asking “Who is behind it?”, we should ask “What structures amplify analysis into polarization?”

That is the posture of a mature democratic society in an era of information warfare.

8. What We Should Truly Be Wary Of

The essence of information warfare is not the border.

Even if there is an external trigger, whether it gains real influence depends on a society’s internal structure.

What we should truly be wary of are structural conditions such as these:

• Algorithmic structures that distort analysis
• Political rhetoric that intensifies friend–enemy binary thinking
• Narratives that amplify distrust in institutions and elections
• The formation of a climate that excludes dissent
• Fear-driven narratives that mobilize emotion

When these take root in society, external influence operations require far less effort to succeed.

In a society with deep polarization, even a small spark can spread into a large fire.

Conversely, in a society where institutional trust remains intact and pluralistic debate is protected, external triggers are far less likely to produce major effects.

The problem is not “which country is the enemy.”

The real problem is how far our society has drifted into binary thinking, and how easily it can be mobilized through fear.

When discussing information operations, it is necessary to be alert to external actors.

But what matters even more is how we constrain domestic amplification structures and how we preserve a free space for public debate.

If those foundations collapse, external intervention is no longer the “cause”— it becomes merely the “catalyst.”

9. What Should We Call an “Information Operation”? (Definitions and Criteria)

As we have seen, the mere existence of foreign-origin posts does not automatically mean they constitute an “information operation.”

So what criteria should we use to determine when something should be called one?

① Intent

Is the activity designed not as ordinary opinion expression, but as an operation engineered to achieve a specific political objective or produce a targeted social effect?

Accidental virality and organically emerging debate should be distinguished from planned influence activity.

② Organization and Coordination

Are multiple accounts acting simultaneously, amplifying the same hashtags, texts, or images with a division of roles?

Concentrated account creation dates, consistent posting patterns, and skewed amplification networks can indicate organizational coordination.

③ Deception

Is the origin or location of the sender deliberately concealed, or is there impersonation designed to mislead domestic users?

Transparent expression should be evaluated by a different standard from inauthentic activity that relies on concealment.

④ Falsity and Misleading Content (Disinformation)

Does it include fabricated facts, manipulated materials, or distortions created through selective editing?

Fact-based criticism and the dissemination of falsehoods must be clearly distinguished.

⑤ Scale and Impact

What was the scale of activity, and what level of impact did it have?

Hundreds of accounts and a few hundred engagements have a different social effect from thousands of accounts and millions of engagements.

Only when these elements overlap does it become reasonable to label something an “information operation.”

What matters is not the nationality of the sender, but the nature of the behavior.

Foreign-origin content by itself does not automatically imply an operation. Likewise, criticism of those in power is not, by itself, evidence of an operation.

If standards remain ambiguous and we begin labeling “unwelcome speech” as foreign operations, the boundary of verification collapses.

That does not strengthen security. It shrinks the space for public debate.

What the information age demands is not simplification driven by fear, but calm judgment grounded in definitions and criteria.

Conclusion

Russia, the United States, and China each shape the information space in different ways.

But for Japan, the central question is not only that.

And one more question.

Information operations exist in reality. That is why verification is necessary.

But once we retreat into the story of “who is behind it,” we end up strengthening a different narrative.

A society that consumes analysis is already leaning—before it is manipulated.

As I read this article, I felt a strong sense of discomfort.

Why is it that criticism of a particular figure is so quickly tied to “foreign influence”?

Why does the weight of the word “operation” change depending on who is being criticized?

And to be completely honest, I also feel a measure of fear.

If an atmosphere spreads in which criticism is reduced to “external threats,” then society may decay from within before it is ever “destroyed from the outside.”

In an era of information warfare, what is most fragile is not the border— but a society’s capacity for reason.