The expert being promoted has not necessarily been invented. They may have written a book, worked in an institution, lived in a country for many years, or developed real expertise on a subject. They may also sincerely believe everything they say. The operation then lies elsewhere: in the reasons their voice is chosen, the platform they receive, and the authority attributed to them. A real person, authentic experience, and a longstanding conviction can become part of an arrangement whose purpose extends far beyond what that person believes they are doing.
This is an essential dimension of manufactured credibility. It is easy to picture an imaginary researcher, a fabricated biography, or a research center without researchers. Such constructions exist, but they cover only part of the subject. Another approach presents a selected opinion as an outside assessment, even though it was chosen precisely because it supports a preferred position. The conclusion was known, or largely predictable, before its author was presented as the person to consult. The public sees expertise that appears to justify the message. It does not necessarily see the selection that placed that expertise before it.
A person can speak sincerely while their promotion serves a strategic calculation. A front allows a position that serves particular interests to be received as an independent judgment. It changes the status the public assigns to the message and the reasons people give themselves for listening. To understand this operation, we begin with documented cases and research on credibility, then examine the connections between the person selected, the organization presenting them, and those who give them credence.
Credibility depends on who is listening
Credibility is relational: someone recognizes a person as competent on a subject under particular circumstances. A title does not have the same effect everywhere. A prestigious affiliation can reassure, leave people indifferent, or evoke an institution already viewed with suspicion. Practical experience can be decisive in understanding a specific problem and insufficient to resolve a different question. Audiences bring their knowledge of the subject, their habits, and their existing relationships of trust to the encounter. Authority also emerges from that encounter.
The strategic value of a voice therefore depends on the audience it addresses and the effect expected of it. An operation may seek recognition for an interpretation, establish someone as an authoritative source, or introduce a question into a debate from which it was previously absent. These objectives describe different outcomes. A voice may be visible to many people yet count for nothing in any decision; it may also circulate on a modest scale within a setting where the subject is actually being discussed. When we speak of an operation’s “target,” the word must remain connected to an intended effect and a context. Reducing it to a mass of people exposed to messages obscures the central question: what place is this voice supposed to occupy in their judgment?
A study by Dan Kahan, Hank Jenkins-Smith, and Donald Braman offers a specific insight. American adults evaluated fictional authors with prestigious academic credentials. The position attributed to those authors varied. Whether participants recognized them as competent, trustworthy experts depended in part on how closely that position aligned with the participants’ orientations. The experiment covered three controversial subjects; it shows that identical credentials can be received differently, without providing a universal model of reception. [1]
This leads to a distinction easily lost when we speak too quickly of “the expert.” There is what the person knows, how they are presented, and the competence their audience recognizes in them. These three elements can coincide. They can also diverge considerably. One reader may recognize authority in the precision of an argument, another in lived experience, and another in intellectual affinity. Nor must the same reader maintain the same criterion from one subject to another. They may demand an explicit method on a question they know well and rely more heavily on reputation when they are less familiar with the subject. Analysis needs to take this variation seriously.
The way someone speaks forms part of this relationship. Vocabulary, certainty, examples, and the space given to objections all contribute to how the speaker is perceived. A highly confident statement may be received as a sign of competence or as an unsupported assertion; technical language can clarify an argument or make it difficult to examine. The form of a contribution tells a story about its author, but that story also depends on who is listening. There is therefore no single appearance of expertise that works mechanically across all audiences.
This cognitive dimension matters. It nevertheless sits within a social history: experiences of institutions, previous conflicts, the interests at stake, and the reasons people have for trusting or distrusting one another.
Choosing a voice whose conclusion already fits
The case of a person selected because their views point in the desired direction deserves a central place. Their conviction may predate by years their relationship with those promoting them. They have their own arguments, references, and sometimes a substantial body of work. The sponsor is therefore not necessarily the origin of their opinion. Its intervention concerns the conditions under which that opinion becomes visible, reaches other interlocutors, and acquires public authority. In this configuration, the analytical question shifts: why was this voice given this position, and what picture of the debate emerges from that choice?
There may be a lasting convergence, agreement confined to one question, or a shared interpretation of an event. These situations do not imply that the person shares every objective of those supporting them. They may defend a position because they believe it is right, because it reflects their experience, or because it follows from a political conviction. Those giving them visibility may, for their part, find an entirely different use for it. The relationship thus allows two logics to coexist: the author believes they are contributing to a debate they understand; the organization sees that contribution as confirmation favorable to its interests. This difference in perspective explains why a person’s sincerity alone cannot account for the arrangement through which their voice circulates.
The conviction is their own. The conditions of their visibility may serve other interests, even when nobody dictates their sentences and no false biography has been created.
Selection can also change the apparent state of available expertise. A debate contains disagreements, uncertainties, and competence distributed across several fields. If the most visible voices converge on the same conclusion because that convergence influenced their promotion, an observer may mistake the result of a selection for the overall state of knowledge. The issue is not that these people are necessarily incompetent. It concerns how representative they are assumed to be. A few authentic opinions can be presented as the expression of a much broader community. A selected opinion then becomes “what the experts think,” even though its quality alone provides no basis for that generalization.
The expertise may even be perfectly suited to the subject. The distortion then concerns the place given to that opinion, its stated independence, or the representativeness attributed to it. Sound expertise is not enough to guarantee the accuracy of these three presentations.
The NATO Strategic Communications Centre of Excellence report Information Laundering in Germany describes, among other practices, the amplification of real political and public figures to lend credibility to the content studied. This finding supports an important distinction: the strategic use of existing authority is a subject of investigation in its own right, separate from the invention of an identity. [5]
It is also necessary to examine what a person’s presence is made to signify. Are they being asked to explain a body of knowledge, invited to describe an experience, or presented as endorsing the whole undertaking? These functions invoke different kinds of authority. A limited contribution can be turned, in its public presentation, into a general endorsement. The person believes they are answering a question; the document surrounding their contribution assigns them a broader position. Part of the operation then lies in the role attributed to them, even if their words have been reproduced accurately.
This does not make every invitation influenced by a preferred viewpoint a covert operation. Media outlets choose their interlocutors, associations defend causes, and institutions support research. The distinction rests on evidence: any deception about the relationship, the control actually exercised, the guarantees of independence claimed, or the misleading presentation of a set of opinions. Convergence, support, and subordination describe different relationships. Conflating them would give the sponsor a degree of control over people that the evidence does not establish.
A book or a title can become a much broader source of authority
The word “expert,” used on its own, explains very little. An expert on what? On the basis of which experience? Using what method, and within what limits? This does not mean expertise is meaningless. Real competence helps us understand problems we would otherwise grasp less well. But the label alone does not tell us where that competence begins and ends. It is precisely in this space that the way a person is presented can produce an unjustified extension of their authority.
A book provides a simple example. It establishes that an author has produced a work whose subject, sources, and quality still need to be assessed. A book devoted to a historical period may reflect thorough research on that period. It does not establish competence across every aspect of the country’s present circumstances. Yet the shift from “the author of a book on this subject” to “the expert on this region” can erase the limits of the original work. The same slippage occurs when a former position becomes a general guarantee of access to current information, or when local experience is treated as knowledge of an entire society. The basis is authentic. The authority inferred from it has changed in scale.
What is manufactured may therefore be the status rather than the person.
That status also derives meaning from the environment in which it is presented. A journalist signals investigative work and editorial responsibility. An association creates an expectation of knowledge of its field or the people it represents. A research center evokes a method, previous work, and institutional continuity. When someone appears in these settings, readers may attribute to them some of the assurances they associate with the organization. The contributor no longer stands alone. They are surrounded by a history, publications, and other names that help readers place their voice. This relationship explains the strategic value of an institution: the recognition granted to the whole can benefit contributions the public has not yet examined in detail.
It remains useful to separate competence from the impression of disinterestedness. In two studies on correcting misinformation, Mark Susmann and Duane Wegener found that expertise could increase the credibility of a correction independently of perceived honesty, particularly when the source had no vested interest in having it accepted. These experiments concern corrections, but they remind us that knowing what one is talking about and being considered trustworthy are distinct dimensions. [12]
A front organization draws on this plurality. It can display expertise, suggest independence, and imply collective recognition. Each claim requires different evidence. A strong contribution does not, by itself, establish the autonomy of its institution; an honorable reputation does not make someone competent on every subject. Apparent authority becomes misleading when these properties are conflated and one is treated as proof of all the others.
What GlobUs and the operations described by OpenAI show
On October 27, 2026, VIGINUM presented its investigations into GlobUs, a non-state structure active particularly in Cameroon and the Republic of the Congo. Described as an experts’ club or a humanitarian association, it reportedly combined a digital presence, public diplomacy, and influence activities in the physical world. The French agency assesses that it serves as a front for information manipulation and interference operations. This attribution is VIGINUM’s. [2]
The AFP dispatch covering the report mentions journalists, influencers, legal specialists, and political scientists, as well as sponsored programs and roundtables. It also reports that the agency considers it likely that the Russian government and its intelligence services use GlobUs, while cautioning against overstating its actual impact on audiences. [3]
The case gives the problem a human and material substance. An organization can have interlocutors, participate in meetings, and leave traces corresponding to real events. People who have encountered it then have direct experience of its existence. They can attest to a discussion, describe a contact, or explain why an activity seemed relevant to them. All of this can be accurate without revealing the relationships that determine the arrangement’s direction. A physical presence therefore does more than provide a backdrop for digital activity. It creates experiences and ties whose public meaning can extend beyond what each participant intended. A photograph establishes that people gathered together; the accompanying account can make it signify agreement, support, or shared views that the image does not demonstrate.
On October 8, 2026, OpenAI described two other configurations. Dark Clark, which the company described as originating in Russia, reportedly controlled the Social Research Center in Latin America, with local contributors who appeared unaware of that control. Bogus Bylines, originating in Iran, used seven deceptive journalistic bylines; nearly one hundred published or republished articles were identified. The company also describes the use of its models for writing tasks and internal reports. [6]
The differences matter. A deceptive byline raises the question of the identity to which the work is attributed. A structure employing real people under concealed control raises the question of how the work is organized. The amplification of an existing public figure raises another question: the selection and reach given to their voice. There is no need to assume a single mechanism behind these situations. They reveal several points at which authority can be constructed or distorted: the person presented, the institution surrounding them, the work associated with them, and the settings in which they are recognized.
Above all, these cases move us beyond the image of an operation composed exclusively of false content and nonexistent people. An arrangement can engage sincere people, publish accurate material, and establish real relationships while misrepresenting its origins or independence. Analysis must be able to hold these elements together. If it sees only what is fictitious, it misses the real substance on which the arrangement rests.
A sincere intermediary retains independent motives and autonomy
In an oversimplified picture of influence, the sponsor thinks, the intermediary executes, and the public receives. This picture leaves out much of the actual work intermediaries perform. A person passes on information because they consider it important, because it relates to a subject they follow, or because it appears to shed light on a situation. They can discuss it, reject it, or interpret it differently. Even when this circulation serves an operator’s interests, it does not necessarily amount to carrying out that operator’s wishes.
Research by Kate Starbird, Ahmer Arif, and Tom Wilson analyzes information operations as a collective activity combining deliberate organization and emergent participation. They emphasize that actors can contribute to the circulation of information without knowing the overall strategy or sharing the operators’ objectives. [4]
This autonomy explains part of a real intermediary’s value, but also the uncertainty they introduce. They bring a reputation and relationships of their own. Their interlocutors know them for reasons that predate the operation. A contribution can therefore attract attention because it fits into that history. At the same time, the person has their own limits: they may refuse an association, correct the way their words have been used, or decide that their supporters have become a liability. The conviction that makes them understandable to their audience can also lead them to oppose the sponsor on another question. Taking intermediaries’ sincerity seriously means recognizing this capacity for disagreement.
The NATO Strategic Communications Centre of Excellence report makes this point regarding Der Tagesspiegel: the misuse of its accurate reporting does not make the outlet an active participant in the operation studied. Work can be used without its author’s knowledge. [5]
There is therefore an analytical cost to describing people too quickly as agents or puppets. We lose sight of their reasons for acting, their disagreements, and the ways they transform what they receive. We also risk overstating the sponsor’s control. A human network cannot be reduced to the list of people from whom an operation has, at some point, derived an advantage.
Several mechanisms can reinforce one another within the same arrangement
Different forms of manufactured credibility can intersect. Someone possesses real but limited expertise; their convictions suit those giving them visibility; an institution presents them with broader authority; their contributions circulate in settings that appear to confirm that recognition. The combination warrants analysis because each element changes how the others are perceived. The institutional reference gives context to the byline, publication leaves a record of reputation, and reputation influences how subsequent publications are read.
The interaction between mechanisms is therefore central. Simply accumulating them does not, however, demonstrate greater effectiveness. An affiliation can strengthen a voice among some interlocutors and cost it credibility elsewhere. Repetition that becomes too conspicuous may arouse suspicion. A recognized contributor may challenge the organization’s presentation of their conclusions. We need to examine the relationships between the elements: what confirms what, what contradicts what, and what changes when several forms of recognition appear together. The explanatory value of their interaction lies in these relationships, rather than in the number of channels or people involved.
Cognitive biases illuminate part of this dynamic. A reputation earned in one field can lead people to attribute much broader competence to someone. Reliance on authority can also push examination of the argument itself into the background. This shortcut must nevertheless be distinguished from reasonable trust in specialists. We all depend on knowledge we cannot verify in full. The problem arises when the assurances relied upon no longer correspond to the question being addressed, or when their presentation misleads the reader.
The confirmation of existing convictions also enters into the evaluation of evidence. In an experiment by Charles Lord, Lee Ross, and Mark Lepper, students examined fictional studies with contradictory findings. They found those supporting their initial position more convincing. The small sample and single issue limit generalization, but the result illuminates a specific point: two people can examine the same evidence using different standards. [11]
Repetition adds another problem. In six experiments, Kimberlee Weaver and her colleagues showed that an opinion repeated by the same person could appear more widely shared. This research concerns the impression of consensus, not the success of a geopolitical operation. It helps explain why familiarity and a plurality of sources need to be distinguished. [7]
These mechanisms can intersect with a very concrete documentary problem. Several texts sometimes cite the same original information. One publication repeats an analysis; another draws on that repetition; a synthesis mentions both. The number of references increases while the number of independent observations remains unchanged. If the reader interprets a diversity of publications as a diversity of evidence, circulation has changed the claim’s apparent status. The information laundering studied by the NATO Strategic Communications Centre of Excellence concerns precisely the legitimization of misleading content through intermediaries that obscure its origins. [5]
Yet it would be reductive to explain everything through a handful of biases that operators supposedly know how to trigger. An audience’s expectations can have historical and material foundations. Distrust of an institution can be grounded in experience. A preference for a witness’s account may reflect an information need that other sources fail to meet. Cognitive mechanisms interact with these reasons; they do not replace them. Our article on hybrid warfare and cognitive biases explores this dimension. Here, it takes a particular form: understanding how a person becomes a recognized authority in a given environment. [13]
Credibility develops within a social environment and is exposed to its judgments
A person’s public standing has a history. It emerges from work, encounters, assessments, and sometimes disagreements that have made a person identifiable within a particular setting. This history explains why a selected expert offers something different from a character invented from scratch. They bring an existence that predates the arrangement, with people who know them and reasons to listen. An influence operation, where one exists, therefore encounters credibility that is already distributed across relationships. It does not create that credibility in its entirety and cannot determine its future alone.
The semi-active networks framework proposed by Hybrid Warfare Insight addresses this human depth: environments that are already active, in which participants pursue their interests and maintain relationships before a crisis gives them new importance. This is an analytical concept developed by the site, and the cases discussed here are not automatically classified within it. [9]
The connection is nevertheless useful. A voice gains weight through more than an institution’s name attached to its byline. It sits within a web of exchanges in which the person’s competence, reliability, and positions are already partially known. This is also where constraints arise. They may have a reputation to protect, colleagues capable of challenging a claim, and an audience that expects a degree of continuity. A new direction may be welcomed, debated, or considered incompatible with what previously gave their work value. Credibility therefore cannot be moved around like a budget its holder could freely allocate to any message.
The tension between control and autonomy runs through the entire arrangement. The sponsor seeks a voice whose effects will favor its interests. Yet that voice’s weight rests partly on the belief that its author could reach a different conclusion. The more a conclusion appears to have been decided before the evidence was examined, the more fragile the appearance of independence becomes. If autonomy is real, the sponsor exposes itself to outcomes it dislikes. This dilemma is particularly visible with people selected because of an initial convergence: agreement on one issue does not imply a willingness to endorse every subsequent position.
Time deepens these relationships in both directions. It gives a reputation depth while leaving more traces on which that reputation may later be reassessed. Older work may be reread in light of a newly revealed affiliation. Participants may clarify the role they believed they were playing. Internal disagreements may become public. Once a concealed relationship has been documented, it changes the context in which previous work is read without automatically making every earlier piece of content false.
Our analysis of the synchronization of instruments in hybrid warfare connects with this point: strategic value depends on their relationships and the effects they produce together. The simultaneous presence of a media outlet, an organization, and public speakers is not enough to demonstrate such integration. We need to understand how their activities intersect and how far they actually reinforce one another. [10]
AI accelerates some tasks without delivering recognition on its own
Artificial intelligence enters this account at a specific level. In the operations described by OpenAI, it was used, among other things, for writing and internal reporting; the company also notes that some operators exaggerated their results. [6]
The ease of producing text that resembles professional analysis changes part of the problem. A polished presentation, fluent language, and a familiar structure become less conclusive when establishing who actually understands the subject. Formal quality can still make a text easier to read, but it guarantees neither the initial observations nor the author’s ability to account for their reasoning. Expertise remains at work when someone asks why a source was selected, what justifies a comparison, or what evidence would change the conclusion. These responsibilities persist regardless of how the sentences were produced.
Producing a document therefore needs to be distinguished from securing recognition for its authority. A tool can intervene directly in the first. The second depends on people, relationships, and institutions with expectations of their own. This is why real intermediaries remain central to this analysis, including when generative tools are involved.
What this authority can actually achieve
Let us return to the objective. An operation may seek acceptance of a claim, but it may also seek recognition for an interlocutor, the repetition of an interpretation, or a shift in the questions an organization must answer. These outcomes are not interchangeable. Someone who is regularly consulted can gain a place in a debate without persuading most of those listening. A publication can prompt a public response without having generated substantial agreement. Evaluation must therefore remain tied to the effect under discussion.
Ben Nimmo’s Breakout Scale allows comparison of operations’ observable spread across platforms, communities, media outlets, and public figures. This reach does not, by itself, demonstrate a change in conviction or a causal political effect. [8]
The distinction also matters within the arrangement. Increased visibility provides evidence that is easy to display: invitations, publications, and citations. These can be presented as signs that the strategy is working, when they primarily establish that activity occurred and a voice circulated. Attributing a broader effect requires understanding what changed among recipients and what evidence connects that change to the operation. An assessment that conflates these levels may satisfy those commissioning it while giving them a false picture of their influence. The biases that affect how messages are received can also affect how organizers interpret their own results.
The issue is therefore more than whether the expert exists, whether their book is real, or whether the media outlet publishing them has an audience. We need to examine the status they have acquired, the assurances attributed to them, and the place their voice actually occupies. This is where the difference emerges between an operation that produces signs of recognition and one to which demonstrated effects can be attributed.
Conclusion
The manufacture of credibility can begin well before the first factual lie. It can emerge in the choice of people promoted, the extension of their expertise, the presentation of their independence, or the impression of consensus created around them. A fictional character is one possibility. A sincere person, sometimes fully competent, selected because their position fits and presented as outside confirmation, is another.
The strategic problem is then to secure recognition for a voice as independent of the interests it serves. Real expertise can give it substance, an institution can provide continuity, and further circulation can bring additional recognition. These elements can reinforce one another, contradict one another, or partly escape those hoping to direct them. Their reception depends on audiences with reasons, experiences, and authorities of their own. Understanding the operation requires looking at this whole while preserving the connection between the message, those carrying it, and the intended effect.
The expert has not always been manufactured. The position they occupy, the independence attributed to them, and the authority derived from their presence may have been.
Sources and related reading
[1] Dan M. Kahan, Hank Jenkins-Smith, and Donald Braman, “Cultural cognition of scientific consensus,” Journal of Risk Research, 14(2), 2011, pp. 147-174. Authors’ version. Read the source.
[2] VIGINUM / SGDSN, “GlobUs : structure d’influence et d’ingérence informationnelle russe en Afrique,” public presentation, October 27, 2026. Read the source.
[3] AFP, “La France dénonce de nouvelles opérations d’ingérence en Afrique au profit de la Russie,” republished by Courrier international, October 27, 2026. Read the source.
[4] Kate Starbird, Ahmer Arif, and Tom Wilson, “Disinformation as Collaborative Work: Surfacing the Participatory Nature of Strategic Information Operations,” Proceedings of the ACM on Human-Computer Interaction, 2019. Authors’ version. Read the source.
[5] Belén Carrasco Rodríguez, Information Laundering in Germany, NATO Strategic Communications Centre of Excellence, 2020. Read the source.
[6] OpenAI, “Disrupting AI-enabled ‘false front’ operations,” October 8, 2026. Read the source.
[7] Kimberlee Weaver, Stephen M. Garcia, Norbert Schwarz, and Dale T. Miller, “Inferring the Popularity of an Opinion From Its Familiarity: A Repetitive Voice Can Sound Like a Chorus,” Journal of Personality and Social Psychology, 92(5), 2007, pp. 821-833. Read the paper (author-hosted PDF).
[8] Ben Nimmo, The Breakout Scale: Measuring the Impact of Influence Operations, Brookings, September 2020. Read the source.
[9] The Analyst, “Semi-Active Networks: The Human Depth of Hybrid Warfare,” Hybrid Warfare Insight, July 29, 2026. Read the source.
[10] The Analyst, “Today’s Hybrid Warfare Is Dangerous, but Still Primitive,” Hybrid Warfare Insight, May 29, 2026. Read the source.
[11] Charles G. Lord, Lee Ross, and Mark R. Lepper, “Biased Assimilation and Attitude Polarization: The Effects of Prior Theories on Subsequently Considered Evidence,” Journal of Personality and Social Psychology, 37(11), 1979, pp. 2098-2109. Read the source.
[12] Mark W. Susmann and Duane T. Wegener, “The independent effects of source expertise and trustworthiness on retraction believability: The moderating role of vested interest,” Memory & Cognition, 51, 2023, pp. 845-861. Read the source. Read the full text on PubMed Central.
[13] The Analyst, “How Hybrid Warfare Exploits Cognitive Biases,” Hybrid Warfare Insight, May 19, 2026. Read the source.