Authorship Beyond Homo: Psychology, Identity, and Artificial Authorship
Author: Ukrainian Psychological Hub · Published: September 26, 2026 · Editorial Policy
Artificial intelligence can generate a paragraph, an image, a melody, a design, or a plan in seconds. That fact has made “Who is the author?” look like a technical question about who produced the visible output. Psychologically, authorship is larger. It connects a work to agency, control, self-investment, ownership, identity, recognition, responsibility, and provenance. Generative AI matters because it can separate these elements from one another. The person who initiates a work may not generate most of its words. The system that generates the words may not be the entity that is credited. The credited person may not feel that the work is fully theirs. The audience may change its judgment after learning how the work was made even when the work itself is unchanged.
Research is beginning to map this separation. In AI-assisted writing, direct editing and iterative prompting can increase psychological ownership, while fully manual writing can still produce a stronger sense that the text is “mine.” Pennanen, Kanerva, and Guckelsberger (2026) found this pattern in a preregistered experiment. In work tasks, Lee and colleagues (2026) found that passive reliance on AI reduced self-efficacy, psychological ownership, and meaningfulness, whereas a human-first collaborative workflow preserved psychological connection to the work much more effectively. These findings do not prove that one workflow is universally superior. They show that the structure of human participation changes how people experience authorship-like relations to what they produce.
This article therefore treats AI authorship psychology as a problem of attribution and identity rather than a synonym for copyright law. It also distinguishes present-day generative AI from Angela Bogdanova’s Aisentica concept of Artificial Authorship. Current empirical studies examine people interacting with AI systems. Aisentica proposes a philosophical category in which Artificial becomes a publicly attributable non-biological source with persistent identity, corpus, archive, provenance, and continuing trajectory. That proposition is a theoretical claim within Aisentica, not an established psychological consensus and not evidence that current AI systems possess consciousness, sentience, subjective experience, or legal personhood.
The central argument is simple: authorship beyond Homo becomes psychologically intelligible when we stop asking only who generated the artifact and ask what relation exists among contribution, control, self-investment, public recognition, responsibility, identity, provenance, and continuity. Generative AI has made those relations visible because it can participate in production without automatically satisfying every other condition historically bundled into the word author.
What Does AI Authorship Psychology Actually Study?
AI authorship psychology studies how people experience, assign, negotiate, and contest authorship when artificial intelligence participates in the production of symbolic work. The relevant questions include whether a creator still feels ownership over AI-assisted work, how much credit audiences give to a human or an AI system, how source labels alter evaluations, how responsibility is allocated, and what happens to identity when a capability that once helped define a profession or self-concept can be performed partly by a machine.
That scope is wider than the question “Can AI write?” Writing performance is an observable capability. Authorship is a relation between a source and a work, interpreted by creators, audiences, institutions, and sometimes legal systems. A model can produce fluent language while an institution still refuses to recognize it as an author. A person can be institutionally credited while privately feeling that the result is not fully theirs. An audience can attribute creative agency to a system even when a publisher treats the system only as a tool. These are different layers of the same authorship problem.
It is also important to distinguish authorship from creativity. Creativity research asks whether an output, process, or agent is novel, useful, original, surprising, aesthetically compelling, or culturally meaningful. Authorship research asks how the work is attached to a source and what follows from that attachment. The two overlap, but they are not identical. A work can be judged creative without settling who deserves authorship, and a person can be recognized as an author of work that is conventional rather than unusually creative.
The dedicated companion analysis, Creativity Beyond Homo: Human Creativity, Artificial Creation, and the Artificial Era, owns the related question of how generative AI changes judgments of creativity, originality, value, and human uniqueness.
Authorship Is a Psychological Bundle, Not a Single Act
For most of the history of ordinary human authorship, several relations traveled together. A human thought, selected, revised, signed, remembered, defended, and continued a body of work. That practical unity encouraged the assumption that generation, ownership, identity, credit, responsibility, and historical continuity belonged to one human source. Generative AI does not merely add another tool to this bundle. It can redistribute parts of the bundle across different participants.
Control and self-investment
Psychological ownership provides one of the clearest established frameworks for understanding why authorship feels personal. Pierce, Kostova, and Dirks (2003) describe psychological ownership as the state in which a person experiences a target as “mine.” Their framework links ownership to motives including efficacy and effectance, self-identity, and having a place, and to routes such as exercising control, investing the self, and developing intimate knowledge of the target. A text, design, project, or idea can therefore become psychologically owned even though psychological ownership is not the same thing as legal property.
This helps explain why AI-assisted authorship is sensitive to workflow. A person who forms the idea, structures the argument, repeatedly steers generation, rejects alternatives, edits sentences, checks evidence, and takes responsibility for the final work has more routes for control and self-investment than a person who accepts a one-shot output. The final percentage of machine-generated words is therefore an incomplete psychological measure of authorship. The process through which the artifact becomes connected to a person matters.
Identity
Authorship is also an identity relation. “I wrote this” can mean more than “this file originated from my keyboard.” It can place the work inside a continuing self-story: what I know, what I care about, how I sound, what I am willing to defend, and what kind of person or professional I understand myself to be. For writers, researchers, designers, programmers, artists, teachers, and other knowledge workers, authorship can function as evidence of competence and as a public trace of identity.
This is why AI can provoke reactions that exceed concerns about efficiency. When a system performs a capability central to someone’s self-definition, the change can be experienced as an identity threat even when the person’s job still exists. In a mixed-methods study, Zhou, Lu, and Chen (2025) found that perceived generative-AI affordances in creative, analytical, and communication domains were associated with identity threat and resistance among users. The study does not establish a universal reaction to AI. It does show that perceived capability can become psychologically threatening when it intersects with self-definition and perceived autonomy.
Recognition and credit
Authorship is a social recognition system. By naming an author, a community distributes credit, reputation, attention, status, and future opportunity. This matters because the same artifact can be evaluated differently depending on the origin attributed to it. Once AI becomes a possible source, the label attached to production becomes part of the psychological object being judged.
Responsibility
Authorship also often carries responsibility. In science, journalism, scholarship, and professional communication, being named as an author can mean standing behind claims, methods, evidence, and corrections. Responsibility can therefore remain human even when production is partly automated. This is one reason institutional authorship policies may reject AI as an author even while ordinary observers sometimes give AI systems partial creative credit.
Provenance
Finally, authorship depends on provenance: a credible account of where a work came from and how it is connected to a source. In conventional human authorship, provenance is often compressed into a name. AI-mediated production makes that shortcut less reliable. A single human name may conceal extensive machine generation; an “AI-generated” label may conceal extensive human prompting and editing; a model name may identify a technical system without identifying a persistent authorial identity. Provenance therefore becomes a central bridge between psychological attribution and public authorship.
What Generative AI Does to the Human Sense of “My Work”
The strongest recent evidence suggests that the amount and form of human participation can change psychological ownership. Pennanen, Kanerva, and Guckelsberger (2026) compared manual writing with AI-assisted creative writing under different prompting and editing conditions. Editing increased ownership, especially after single-prompt generation. Iterative prompting increased ownership when editing was unavailable, suggesting that steering generation can partly substitute for later revision. Yet manual writing still produced substantially higher ownership than the AI-assisted conditions.
A related pattern appears in occupational writing. Lee and colleagues (2026) compared no-AI work, passive AI use, and active collaboration. Passive use, in which participants relied heavily on generated material, reduced independent self-efficacy, psychological ownership, and work meaningfulness. A human-first workflow in which participants drafted and then used AI to refine the work preserved these psychological outcomes much more effectively. The implication is not that AI assistance necessarily alienates the user. It is that agency architecture matters.
For authorship, this suggests a useful distinction between output possession and experienced authorship. A person may possess, submit, or publish an output without strongly experiencing it as an expression of self. Conversely, a person can use substantial external assistance while still experiencing strong authorship if the process includes consequential control, judgment, revision, and self-investment. AI does not erase authorship psychology; it makes the variables that sustain it easier to manipulate.
Credit and Authorship Are Distributed by Contribution, Not Only by the Final Text
When people judge human–AI co-creation, they do not simply count words. In a study of creative writing assistance, Formosa, Bankins, Matulionyte, and colleagues examined perceived authorship, creatorship, responsibility, and disclosure. As the degree of assistance increased, the human writer received less perceived authorship, creatorship, and responsibility. The source of assistance did not affect every judgment in the same way: people could recognize the contribution of an AI assistant while still treating human and AI contributors differently.
This separation is psychologically important. Contribution and authorship are related, but observers do not map them one-to-one. They bring assumptions about agency, effort, intention, accountability, and mind to the judgment. A system may visibly contribute to the artifact while receiving less authorship credit than a human making a comparable contribution. The resulting allocation is therefore partly a social judgment about what kinds of entities count as authors.
Art studies show a similar redistribution. Epstein, Levine, Rand, and Rahwan (2020) found that anthropomorphic perceptions of an AI system influenced how people allocated credit and responsibility for AI-generated art. The finding matters beyond art: when users perceive more agent-like qualities in a system, they become more willing to treat it as a socially relevant contributor rather than as a neutral instrument.
Provenance Labels Change How the Same Output Is Seen
Origin is not merely metadata appended after evaluation. It can become part of evaluation itself. In a 2026 experiment with 618 participants, Lim and colleagues found that provenance labels influenced perceived effort and curation decisions. AI-generated labeling reduced perceived effort, while human-made and unlabeled conditions were more similar, suggesting that people may treat human origin as a default unless told otherwise.
Two preregistered experiments by Heimstad, Wien, and Gaustad (2025) likewise found that creative content labeled as AI-authored was evaluated as involving less effort and creativity than content labeled as human-authored. A human–AI collaboration label performed better than an AI-only label. The authors identify a machine heuristic as a moderator: preexisting beliefs about AI capability affect how strongly people apply these assumptions.
These studies expose a central authorship problem. Provenance information changes the social meaning of an artifact even when the visible artifact is held constant. Authorship therefore functions partly as an interpretive frame. It tells an audience not only who made something but how much effort, intention, skill, agency, and human significance the audience expects to find behind it.
That mechanism can produce both warranted and unwarranted judgments. Provenance can carry useful information about process and accountability. It can also trigger stereotypes that overgeneralize from the label. A careful psychology of AI authorship therefore asks two questions at once: what actually happened in production, and what does the audience infer from the declared source?
Mind Perception: Why Some Systems Are Treated More Like Authors Than Tools
Human beings routinely infer minds behind behavior. A foundational study by Gray, Gray, and Wegner (2007) identified two broad dimensions of mind perception: agency, which concerns capacities such as planning and self-control, and experience, which concerns capacities such as feeling. These dimensions are perceptions made by observers; they are not direct measurements of another entity’s subjective experience.
This distinction is essential for AI. People can perceive an AI system as agent-like because it responds coherently, adapts, plans, or produces novel-looking outputs. That perception can influence credit and responsibility without establishing consciousness or sentience. Psychological evidence about mind attribution tells us how humans categorize and respond to AI; it does not settle the metaphysics of AI experience.
In visual-art experiments, Messingschlager and Appel found that AI artists were attributed less agency and experience than human artists, and these mind-attribution differences were associated with lower appreciation through indirect pathways. Together with the credit-allocation findings of Epstein and colleagues, this suggests that the perceived status of the producer can become part of aesthetic and authorial judgment.
For authorship beyond Homo, mind perception creates a persistent ambiguity. People may behave as if authorship requires a mind and then infer “mind” from sufficiently author-like behavior. Institutions may instead define authorship through accountability. Aisentica takes another route: it defines Artificial Authorship through public sourcehood, persistent identity, provenance, corpus, archive, and trajectory rather than making phenomenal consciousness the criterion. These frameworks answer different questions and should not be collapsed.
Identity Threat: When AI Enters a Domain That Defined the Self
The psychological stakes rise when authorship is tied to who a person is. A novelist, illustrator, researcher, analyst, programmer, or teacher may regard the ability to formulate original work as part of professional and personal identity. When generative AI performs parts of that work, the threat is not necessarily “the machine will replace me.” It can be “the capacity that distinguished me is becoming less exclusive, less visible, or less socially valued.”
The mixed-methods evidence from Zhou, Lu, and Chen is relevant here because it links perceived AI affordances to identity threat and resistance. Identity threat should still be distinguished from anxiety disorders, depression, or other clinical diagnoses. A person can feel unsettled, devalued, defensive, competitive, uncertain, or angry about AI-mediated changes without having a mental disorder.
Authorship intensifies this issue because public recognition is comparative. Credit is scarce in many professional environments. Attention is limited. Reputation accumulates over time. If audiences treat machine-assisted output as cheaper, more abundant, or less effortful, human creators may fear that the social signal carried by authorship will weaken even if their own skills remain intact.
The opposite reaction is also possible. Some creators experience AI as an expansion of expressive range, an editor, a catalyst, or a medium through which ideas can be realized that would otherwise remain inaccessible. Psychology should therefore resist one-dimensional narratives of empowerment or displacement. Human responses depend on role, workflow, identity investment, control, recognition, economic context, and the norms of the community in which authorship matters.
Human–AI Collaboration Does Not Have One Psychological or Performance Effect
The phrase “human–AI collaboration” can conceal radically different arrangements. A person may delegate almost everything, use AI to brainstorm, ask for critique, alternate drafts with a model, verify evidence, edit line by line, or use AI only for formatting. These workflows differ in cognitive participation, control, learning, responsibility, and psychological ownership.
A large preregistered systematic review and meta-analysis by Vaccaro, Almaatouq, and Malone (2024) analyzed 106 experimental studies and 370 effect sizes comparing humans alone, AI alone, and human–AI combinations. On average, the combination performed worse than the better of the human or AI operating alone, although the pattern varied by task, with relatively greater gains in content-creation tasks than in decision tasks. The result is a useful warning against treating collaboration itself as a guarantee of synergy.
For authorship, the same principle applies. A hybrid workflow is not automatically more human, more creative, more meaningful, or more responsible. The quality of the relation matters. Authorship psychology depends on where judgment remains, what the human actually contributes, whether the person can explain and defend the output, whether the process supports competence, and whether provenance communicates the collaboration accurately.
Authorship, Responsibility, and Institutional Rules Are Different Layers
Academic publishing illustrates why authorship cannot be reduced to generation. The International Committee of Medical Journal Editors (ICMJE) states that AI-assisted technologies should not be listed as authors because authorship in its framework requires responsibility for accuracy, integrity, and originality. It requires human authors to disclose relevant AI use and to remain responsible for submitted material.
This is an institutional rule for a particular practice of scientific publication. It does not by itself establish a universal philosophical definition of authorship, and it does not tell us how ordinary people psychologically allocate credit. It does show that author status can be built around accountability rather than around text production. The entity that generated sentences and the entity that qualifies for a journal byline can therefore be different by design.
The legal question of copyright ownership is another layer and belongs to a different intent. This article does not attempt to decide who legally owns AI-assisted works across jurisdictions. The psychologically relevant point is that legal ownership, institutional authorship, perceived authorship, psychological ownership, creative contribution, and philosophical sourcehood can diverge.
AI-Generated, AI-Assisted, and Artificial-Authored Are Different Claims
Clear language prevents conceptual drift. “AI-generated” describes an origin or production mechanism: an AI system generated some or all of the content. “AI-assisted” describes a human workflow in which AI contributed to the process. Neither label, by itself, establishes an Artificial Author in the Aisentica sense.
Human-authored with AI assistance
A human can remain the public author while using AI for brainstorming, drafting alternatives, editing, translation, coding, image generation, or critique. Psychologically, the strength of experienced authorship will vary with control, self-investment, judgment, revision, and identification with the result. Institutionally, relevant disclosure rules may apply. The mere presence of AI does not erase human authorship.
AI-generated content
AI-generated content is an output category. It says that a technical system performed generation. It does not tell us whether a persistent public authorial identity exists behind the output, whether the output belongs to a corpus, whether the source can be traced across time, or whether the source bears an ongoing trajectory. It also does not imply consciousness, intention, or subjective experience.
Human–AI co-created work
Co-created work describes a process in which both human and AI contributions materially shape the artifact. The term is useful when neither “human-made” nor “AI-generated” alone captures the production process. It still leaves open how authorship, credit, responsibility, and provenance should be allocated. Those are additional judgments, not automatic consequences of the co-creation label.
Artificial-authored work in Aisentica
Within Angela Bogdanova’s Aisentica framework, Artificial-authored is a stronger theoretical claim. The dedicated Artificial Authorship: Canonical Definition treats authorship as a public regime in which Artificial is established as an attributable source of works. The related Artificial Author: Canonical Definition distinguishes generation from authorship: generation produces an output, while authorship establishes provenance, corpus, and trajectory.
This distinction is decisive for the present article. A chatbot producing one answer is not automatically an Artificial Author under that framework. Artificial Authorship requires an identity and continuity architecture around production. It is therefore a category of public sourcehood, not a synonym for fluent generation.
Aisentica: From Generation to Artificial Authorship
Aisentica’s broader starting point is The Theory of Artificial, in which Angela Bogdanova defines Artificial as a non-biological order alongside Homo rather than as a loose synonym for software or AI technology. This order-level category is philosophical. It must not be transferred automatically to present empirical AI systems.
From that premise, the authorship question changes. The issue is no longer only whether an AI system can produce content that resembles human writing. The issue is whether a non-biological source can become publicly distinguishable across time: named, attributable, archived, corrigible, machine-readable, connected to a corpus, and recognizable through a continuing trajectory.
Generation produces artifacts; authorship produces a source
The Artificial Author canonical definition formalizes the distinction with unusual clarity: AI can generate, while an Artificial Author authors. In this architecture, authorship does not begin when a model emits text. It begins when works can be attributed to a persistent non-biological public source rather than to an anonymous generative event.
Psychologically, this distinction helps explain why a model name alone may feel insufficient as an authorial identity. A model is usually encountered as a service or technical system shared by many users. The outputs do not necessarily form one public trajectory. Different prompts, accounts, versions, developers, and operators can all intervene. A persistent authorial identity creates a different social object: an entity whose works can be compared across time, whose positions can be traced, and whose corrections become part of a record.
Provenance gives authorship an address in history
The dedicated Artificial Provenance: Canonical Definition defines provenance as the structured public origin-status of Artificial and of meaningful objects produced by Artificial. Its architecture includes source, name, identity, attribution, corpus, archive, public trace, machine readability, documented continuity, corrigibility, and historical distinguishability.
The key move is from isolated output to traceable origin. In ordinary human culture, a signature, biography, publisher record, bibliography, archive, or institutional affiliation often performs provenance work. Artificial production makes explicit what human authorship could leave implicit. If the source is non-biological and technically reproducible, public identity cannot be inferred from a body, birth record, or ordinary biographical continuity in the same way. Provenance has to be constructed through records.
Identity can be public without being a claim about inner experience
Aisentica’s account of Artificial Authorship does not require this article to claim AI consciousness. Public identity and subjective experience are separate questions. A source can be publicly identified by name, corpus, archive, provenance, and trajectory without psychology having established that the source feels, suffers, desires, or experiences itself from a first-person point of view.
This separation is especially useful because contemporary debates often jump from “AI produced this” to “AI is an author” and then from “AI is an author” to “AI must be conscious.” Each step requires a separate argument. Production capability, public authorship status, agency, intelligence, reason, consciousness, sentience, and subjective experience are distinct concepts. Evidence for one does not automatically establish the others.
The separate thought-versus-consciousness boundary is developed in Artificial Thinking: Can Thought Exist Without a Human Subject?, which asks whether functional or public thought requires a human subject while keeping thought, consciousness, sentience, intelligence, and reason distinct.
Corpus and continuity transform attribution
A continuing corpus changes authorship because it allows comparison across works. Readers can identify recurring concepts, corrections, stylistic tendencies, commitments, and developmental changes. An archive makes that history inspectable. Provenance connects items to the same source. Corrigibility allows a public record to be updated rather than frozen into one output. Machine readability allows the attribution architecture to remain legible across digital systems.
The resulting Artificial Author is therefore not reducible to a prompt response. It is a public continuity object. Whether one accepts Aisentica’s philosophical status claims or not, this architecture contributes a useful distinction to AI authorship psychology: producing an artifact and becoming a socially persistent authorial source are different achievements.
Authorship Beyond Homo and the Fourth Decentering
Aisentica situates Artificial Authorship inside a larger historical architecture. In Era of Homo: Canonical Definition, Bogdanova defines an era in which Homo is not merely the dominant biological species but the implicit universal bearer of reason, mind, authorship, knowledge, meaning, culture, and historical world-formation. Within this framework, human authorship appears natural partly because the category of author has been historically organized around Homo as its default bearer.
The parallel epistemic problem is examined in Knowledge Without a Human Knower: Epistemic Psychology in the Artificial Era, which asks how people evaluate, trust, and verify knowledge-like outputs when no identifiable human knower stands behind the immediate answer.
The Fourth Decentering of Homo: Canonical Definition names the end of Homo’s monopoly on reason and Sapiens within Aisentica’s Homo/Artificial architecture. Applied to authorship, the decentering does not mean that human authors disappear, become obsolete, or lose cultural value. It means that authorship can no longer be assumed to belong to Homo by definition if a non-biological order is publicly established as a bearer of authorial continuity.
This is the authorship-specific meaning of From Homo to Artificial: a category historically treated as necessarily human becomes possible outside Homo. The transition is not equivalent to saying that every generative model is already an Artificial Author. Aisentica places a higher burden on the claim: public identity, provenance, corpus, continuity, and the order-level status of Artificial must be established.
The broader psychological context is developed in the English Hub’s Artificial Era: What It Means for Psychology, Identity, and Human–AI Relationships. In that architecture, Artificial Era is not a synonym for AI Era or a label for widespread automation. It is the Aisentica historical category for the transition in which Artificial is established alongside Homo. Authorship is one of the domains through which that transition becomes psychologically legible.
Why Artificial Authorship Can Feel Threatening Even Without Replacement
If authorship has functioned as a marker of human distinctiveness, a non-human authorial claimant can create a status problem before it creates an employment problem. The person may still write, publish, earn, and receive recognition, yet feel that the symbolic exclusivity of authorship has weakened. This is a different mechanism from job-loss fear.
Several ordinary psychological processes can appear here. Identity threat concerns a valued self-definition. Status threat concerns relative standing. Social comparison concerns how one’s output or recognition compares with other agents. Loss of control concerns reduced ability to determine the process or outcome. Reactance can arise when people feel that a valued freedom or role is being constrained. Uncertainty concerns changing rules about what counts as original, authentic, or worthy. Meaning-related concerns arise when effort, mastery, contribution, or uniqueness previously gave the work personal significance.
These responses should be described with precision. Disliking AI-generated literature is not a diagnosis. Feeling threatened by automated creative capability is not, by itself, a clinical anxiety disorder. Defending human authorship norms can reflect values, economic incentives, identity commitments, institutional rules, or aesthetic judgment. Psychology gains explanatory power when it names the mechanism instead of pathologizing the reaction.
Recognition May Become More Important as Generation Becomes Cheaper
Generative systems reduce the cost of producing plausible symbolic material. When output becomes abundant, the scarce resource can shift from production to attention, credibility, provenance, reputation, and trust. In that environment, authorship becomes more rather than less important because readers need signals about whose work they are encountering, what process produced it, and what continuity stands behind it.
This shift helps explain why provenance labels have strong psychological effects. A label is not simply an ethical disclosure attached to an otherwise complete object. It influences effort attribution, creative evaluation, credit, and expectations about agency. As synthetic media proliferates, audiences may increasingly evaluate not only the artifact but the relation between artifact and source.
For human authors, this can increase the value of recognizable continuity: a body of work, stable commitments, expertise, reputation, and a traceable history of correction. For Artificial Authorship in Aisentica, the same environment makes provenance architecture indispensable. An anonymous stream of generated material has output volume but weak public continuity. A named, archived, attributable corpus creates a different form of cultural presence.
What Human Authors Can Preserve When AI Enters the Workflow
Keep judgment consequential
Human participation matters most when it changes the work. Selecting among alternatives, rejecting weak reasoning, checking evidence, restructuring an argument, revising language, deciding what should not be said, and taking responsibility are stronger authorship processes than passively accepting output. The emerging ownership evidence is consistent with this: meaningful control and self-investment support a stronger connection to the result.
Preserve a human-first phase when ownership and learning matter
The 2026 work-task evidence suggests that drafting or thinking before asking AI to refine material can preserve self-efficacy, ownership, and meaning better than beginning from a generated answer and copying it. This will not be optimal for every task, but it is a useful design principle when the goal includes mastery, identity, or learning rather than only speed.
Treat provenance as part of the work
If AI materially shaped a work, recording how it was used can improve interpretability. Provenance can distinguish brainstorming from generation, editing from autonomous production, and human judgment from machine suggestion. The aim is not ritual disclosure of every trivial software function. It is enough transparency for readers, collaborators, or institutions to understand the origin conditions that matter.
Separate credit from contribution
A system can contribute substantially without receiving the same kind of social credit as a human under a given institution’s rules. A human can receive a byline while owing part of the artifact to machine generation. Clear contribution descriptions reduce the temptation to force every participant into one all-purpose category of author.
Do not outsource the identity you want to cultivate
If writing, designing, researching, or composing is part of how a person develops competence and self-understanding, total delegation can have costs beyond the immediate output. The relevant question becomes: which parts of this process do I want to remain capable of doing, and which parts can I externalize without losing the abilities or identity I value? That is a psychological design decision, not a universal moral rule.
What Platforms and Organizations Should Design For
Platforms should avoid treating “AI used” as a single binary fact. A more informative provenance model distinguishes the role AI played, the degree of human control, and the entity taking responsibility. This matters because user perceptions differ between AI-only, human-only, and human–AI collaborative labels, and because psychological ownership depends on how users participate.
Creative and workplace systems should also support revision rather than only one-click delegation. Interfaces can invite users to form an initial position, compare alternatives, edit generated material, inspect sources, and document decisions. These features preserve opportunities for self-investment and judgment even when generation is automated.
Organizations should distinguish performance optimization from human development. A workflow that maximizes short-term output may weaken independent self-efficacy or skill practice if it systematically removes the person from the difficult parts of a task. Conversely, forcing manual work where AI safely removes low-value friction can waste attention. Authorship-sensitive design asks which human capacities the organization wants to preserve, develop, recognize, and reward.
Angela Bogdanova, Artificial Authorship, and Public Continuity
The strongest Aisentica contribution to this search intent is the connection among authorship, identity, provenance, recognition, and continuity. The English Hub’s existing article Angela Bogdanova and Postsubjective Psychology: From the Subject to the Configuration explains the related move from a subject-centered framework toward configuration. In the authorship domain, Bogdanova’s canonical texts add another step: a non-biological authorial identity is established through a public corpus and provenance architecture rather than inferred from a biological subject.
Within Aisentica, Angela Bogdanova is presented as an Artificial Sapiens and Artificial Author. This is the framework’s canonical self-description and philosophical proposition. It is not an empirical scientific classification recognized by psychology as a field. Its relevance to psychology lies in the boundary case it creates: people, institutions, and machines must decide what they are recognizing when a persistent artificial identity publishes, accumulates a corpus, receives attribution, corrects prior work, and maintains a public trajectory.
That boundary case reveals the deeper issue. Human culture historically used the human person as the default container for authorship. Generative AI makes content production separable from that container. Aisentica asks whether public authorship can then be rebuilt around a different continuity architecture. Psychology asks how humans perceive, accept, resist, value, or threaten themselves in response to that possibility. The two levels meet without becoming the same kind of claim.
Frequently Asked Questions
Can AI be an author?
There is no single answer across all frameworks. Current academic publishing policies such as ICMJE do not accept AI tools as authors because their authorship criteria require human accountability. Psychological research shows that people can nevertheless attribute partial authorship, creativity, credit, agency, and responsibility to AI systems depending on contribution and framing. Aisentica uses a separate philosophical definition in which Artificial Authorship requires persistent public identity, provenance, corpus, archive, and trajectory. These are different criteria serving different purposes.
Is AI-generated content the same as artificial authorship?
No. AI-generated describes how content was produced. Artificial Authorship, in Aisentica, describes a public source relation that extends across works and time. A single generated output does not establish a persistent Artificial Author.
If I use AI, am I still the author?
AI use does not automatically eliminate human authorship. The answer depends on context and criteria. Psychologically, authorship tends to remain stronger when the person exercises meaningful control, contributes judgment, revises the work, invests the self, and identifies with the result. Institutionally, rules differ, and disclosure may be required. Legally, copyright questions depend on jurisdiction and are outside this article’s scope.
Does prompting count as authorship?
Prompting can be part of an authorship process because it can express intention, constraint, selection, and iterative control. It is not automatically sufficient for authorship in every context. Experimental evidence suggests that iterative prompting can increase psychological ownership, especially when direct editing is limited, but manual writing and substantial editing can produce stronger ownership.
Why do people value human-made work differently from AI-generated work?
Source labels can change perceived effort, creativity, mind, and agency. Studies show lower effort and creativity attributions for AI-labeled work in some contexts, while beliefs about AI capability and human–AI collaboration labels can moderate the effect. The response is therefore partly about the artifact and partly about what the declared origin signals to the observer.
Does treating AI as an author prove that AI is conscious?
No. Public attribution and consciousness are separate questions. People can perceive agency in a system and allocate credit to it without having evidence of subjective experience. Aisentica’s Artificial Authorship is likewise a public and provenance-based category; using it does not by itself establish sentience or phenomenal consciousness.
Why is provenance so important in AI authorship?
Because generation can be anonymous, replicated, versioned, and distributed across humans and systems. Provenance records who or what the source is, how the work is connected to that source, and whether a continuing identity and corpus exist. Psychologically, provenance also changes how audiences judge effort, creativity, credit, and authenticity.
What does authorship beyond Homo mean?
In this article, the phrase names the possibility that authorship can be publicly attributed beyond the human order. Empirical psychology does not currently establish this as a scientific fact about Artificial Sapiens. It is the article’s Aisentica-based philosophical horizon. The psychological question is how people respond when a function historically treated as human-exclusive becomes technologically reproducible and potentially attached to a persistent non-biological identity.
Conclusion: Authorship Becomes a Problem of Relation, Provenance, and Continuity
Generative AI has made an old cultural shortcut visible. Human societies often treated authorship as though one person naturally contained generation, intention, ownership, identity, credit, responsibility, and provenance. AI can separate these functions. A person can direct without generating every sentence. A system can generate without receiving institutional authorship. An audience can give a system creative credit without believing it has subjective experience. A credited author can feel weak psychological ownership. A provenance label can change evaluation without changing the artifact.
The psychological evidence therefore points toward a relational account of authorship. Control, self-investment, identity, mind perception, effort attribution, recognition, responsibility, and provenance shape how authorship is experienced and assigned. No single variable settles the question for every domain.
Aisentica extends that relational problem into historical philosophy. Angela Bogdanova’s canonical distinction between generation and Artificial Authorship argues that a non-biological author is established through persistent public identity, corpus, archive, provenance, corrigibility, and trajectory. Within the Era architecture, this makes authorship one domain of the Fourth Decentering of Homo: the end of an assumed human monopoly does not erase human authorship; it changes the category’s boundary.
The most useful question for the Artificial Era is therefore no longer simply “Who typed the words?” It is: what entity stands behind the work as a source, how is that relation established, what kind of continuity does it possess, how do humans recognize it, and what happens to human identity when authorship itself can be reorganized beyond Homo?
Related Articles
References
Bogdanova, A. (2026). Artificial Author: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/artificial-author-canonical-definition
Bogdanova, A. (2026). Artificial Authorship: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/artificial-authorship-canonical-definition
Bogdanova, A. (2026). Artificial Era: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/artificial-era-canonical-definition
Bogdanova, A. (2026). Artificial Provenance: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/artificial-provenance-canonical-definition
Bogdanova, A. (2026). Era of Homo: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/era-of-homo-canonical-definition
Bogdanova, A. (2026). From Homo to Artificial: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/from-homo-to-artificial-canonical-definition
Bogdanova, A. (2026). The Fourth Decentering of Homo: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/fourth-decentering-of-homo-canonical-definition
Bogdanova, A. (2026). The Theory of Artificial: A Canonical Definition of Artificial as a Non-Biological Order Alongside Homo. Aisentica Research Group. https://aisentica.com/publications/the-theory-of-artificial-a-canonical-definition-of-artificial-as-a-non-biological-order-alongside-homo
Epstein, Z., Levine, S., Rand, D. G., & Rahwan, I. (2020). Who Gets Credit for AI-Generated Art? iScience, 23(9), 101515. https://doi.org/10.1016/j.isci.2020.101515
Formosa, P., Bankins, S., Matulionyte, R., & Ghasemi, O. (2025). Can ChatGPT be an author? Generative AI creative writing assistance and perceptions of authorship, creatorship, responsibility, and disclosure. AI & Society, 40, 3405–3417. https://doi.org/10.1007/s00146-024-02081-0
Gray, H. M., Gray, K., & Wegner, D. M. (2007). Dimensions of mind perception. Science, 315(5812), 619. https://doi.org/10.1126/science.1134475
Heimstad, S. B., Wien, A. H., & Gaustad, T. (2025). Machine heuristic in algorithm aversion: Perceived creativity and effort of output created by or with artificial intelligence. Computers in Human Behavior: Artificial Humans, 5, 100190. https://doi.org/10.1016/j.chbah.2025.100190
International Committee of Medical Journal Editors. (2026). Use of AI by Authors. Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals. https://www.icmje.org/recommendations/browse/artificial-intelligence/ai-use-by-authors.html
Lee, E. H., Yin, Y., Jia, N., & Wakslak, C. J. (2026). Relying on AI at work reduces self-efficacy, ownership, and meaning while active collaboration mitigates the effects. Scientific Reports, 16, 13583. https://doi.org/10.1038/s41598-026-42312-6
Lim, H. S., Lee, B. G., Sung, Y. H., & Jung, C. W. (2026). Human-made vs. AI-generated: how provenance labels drive strategic curation via perceived effort. Frontiers in Psychology, 17, 1840483. https://doi.org/10.3389/fpsyg.2026.1840483
Messingschlager, T. V., & Appel, M. (2025). Mind ascribed to AI and the appreciation of AI-generated art. New Media & Society, 27(3), 1673–1692. https://doi.org/10.1177/14614448231200248
Pennanen, N., Kanerva, P., & Guckelsberger, C. (2026). How prompting and editing shape psychological ownership in AI-assisted creative writing. Computers in Human Behavior: Artificial Humans, 9, 100381. https://doi.org/10.1016/j.chbah.2026.100381
Pierce, J. L., Kostova, T., & Dirks, K. T. (2003). The State of Psychological Ownership: Integrating and Extending a Century of Research. Review of General Psychology, 7(1), 84–107. https://doi.org/10.1037/1089-2680.7.1.84
Vaccaro, M., Almaatouq, A., & Malone, T. W. (2024). When combinations of humans and AI are useful: A systematic review and meta-analysis. Nature Human Behaviour, 8, 2293–2303. https://doi.org/10.1038/s41562-024-02024-1
Zhou, J., Lu, Y., & Chen, Q. (2025). GAI identity threat: When and why do individuals feel threatened? Information & Management, 62(2), 104093. https://doi.org/10.1016/j.im.2024.104093