Rethinking the SECI model for GenAI: three perspectives, including one from Japan

The SECI model¹ was first proposed in 1994-1995 by Japanese knowledge management (KM) thought leaders Ikujiro Nonaka and Hirotaka Takeuchi. It remains popular in KM, with some 30 articles in RealKM Magazine exploring on the one hand case studies in which it has been usefully applied, and on the other hand, criticisms including potential alternatives. These criticisms include that SECI is too dated, too simple, and not as widely applicable as often assumed. For example, as Hanna M. Koskinen explores in section 2.5 of her notable 2025 thesis developing the core principles of responsible KM² and discussed in section 3.1 of John Edwards and Antti Lönnqvist’s landmark 2023 review of the future of KM research and practice³ and my comments on that section.

Given the ongoing popularity of the SECI model and the disruptive emergence and rapid uptake of generative AI (GenAI), it’s unsurprising that a growing number of researchers are examining the SECI model in the context of GenAI. This article presents three such research perspectives:

• The GRAI framework proposed by Karsten Böhm and Susanne Durst.
• The hybrid SECI framework proposed by Roberto Cerchione, Giuseppe Liccardo, and Renato Passaro (whose research has also brought us the very valuable GenAI-KM organizational readiness self-assessment tool^4,5).
• The GenAI SECI model proposed by Naoshi Uchihira, with contributing research also involving Takuichi Nishimura and Koki Ijuin.

A fourth perspective also exists which proposes the AKI (artificial knowledge integration) model as a new model inspired by the SECI model. However, I can’t discuss it here as it is presented in a chapter that is not an open knowledge resource, having been published in a paywalled academic book. The use of the AKI acronym is also potentially confusing given that David Williams has already long been using AKI (action-knowledge-information) model⁶ as the name of his alternative approach to the DIKW (data-information-knowledge-wisdom) model, another popular early KM model which is facing strong criticism⁷.

Each of the three perspectives exploring the SECI model in the context of GenAI is elaborated below, following on from an introduction to the SECI model.

But what about the Japanese perspective?

An argument that is increasingly being put forward in defence of the SECI model is that criticisms of SECI are not considering its cultural context. This argument is very valid.

Nonaka and Takeuchi developed the SECI model as a result of their studies on innovation in Japanese companies in the 1980s and 1990s, so the model reflects the values and culture of Japanese business and work practices. These differ significantly from the values and culture of Europe and North America, where much of the power base of KM is located. For example, Japan is a high-context culture^8,9.

Ignoring the cultural context of SECI is an example of the continued coloniality in KM, which as I argue in a recent feature article¹⁰, the KM community needs to move beyond. One of the three perspectives presented below offers us the opportunity to progress this. Just like Ikujiro Nonaka and Hirotaka Takeuchi, who proposed the original SECI model, Naoshi Uchihira, who is proposing the GenAI SECI Model, is Japanese, and so are his collaborators Takuichi Nishimura and Koki Ijuin. All three are researchers with the Japan Advanced Institute of Science and Technology.

What do you notice is different about Naoshi Uchihira’s perspective below?

The SECI model

Before elaborating each of the three perspectives exploring the SECI model in the context of GenAI, a basic introduction to the SECI model is useful for those unfamiliar with it. If you already have a good knowledge of SECI, feel free to skip this section and directly engage with the three perspectives below.

The SECI model describes the processes of knowledge creation in companies and emphasizes the interplay between tacit and explicit knowledge. SECI provides insights into how knowledge can be created, converted and transferred with an iterative approach based on four phases of socialization, externalization, combination, and internalization, hence the acronym SECI. Knowledge creation in the model progresses through a spiral form rather than a circular movement (Figure 1). This is because knowledge moves up from the individual level to the group level and further (and finally) to the organizational level.

The SECI process takes place in “ba” which has been defined as a shared context in which knowledge is shared, created, and utilized. Consequently, context needs to be considered when trying to create meanings.

In the SECI model:

• Socialization (the S in SECI) refers to the exchange of knowledge between human beings.
• Externalization (the E in SECI model) represents the explication of (internalized) knowledge in some form of externalized information (codified knowledge).
• Combination (the C in SECI model) traditionally combines externalized knowledge (e.g. stored information in an IT system).
• Internalization (the I in SECI model) is the process of consuming information (externalized knowledge) into an internal representation that the human user can act upon after it has been internalized.

Perspectives on SECI in the context of GenAI:

1. GRAI framework (Böhm & Durst)

Karsten Böhm and Susanne Durst propose the GRAI (generative, receptive artificial intelligence) framework in a 2025 open access Creative Commons paper¹¹ published in VINE Journal of Information and Knowledge Management Systems. The framework was developed through an envisioning approach based on previous literature and Böhm & Durst’s thoughts and experiences.

The GRAI framework has already been summarised in a previous RealKM Magazine article¹², which at the time of writing this new article had already attracted more that 6,000 views, highlighting the significant level of interest in the topic of SECI in the context of GenAI. The following summary of the GRAI framework is taken largely verbatim from the previous summary article.

Böhm & Durst contend that a revised SECI model should take the machine into account as a participant that can play either an active or a passive role. The active role would generate an output or a response, while the passive role could be compared to listening and adapting/rebuilding the internal (knowledge) representation. Consequently, the four areas would each be split into a human perspective and a machine perspective, leading to eight fields of action in the new GRAI framework (Figure 2).

Sticking to the original knowledge conversion cycle from socialization to externalization to combination and finally to internalization, Böhm & Durst report that GRAI opens up a number of new relationships besides the classical ones that were always assumed to be human-to-human. In the current development stage of generative AI, the most interesting fields are those in which humans and machines interact with each other. This leads to a combination of two actors (human and machine) within two role positions in the four fields of the original SECI model. The resulting eight different interaction fields are summarized below.

Böhm & Durst advise that the situation in which those knowledge exchanges are completely left to the machine(s) can be considered as a topic for future development and investigation, although the first experiments toward this direction are already appearing.

The addition of the machine as an actor in knowledge creation processes does not mean that Böhm & Durst understand both the human and machine roles as equal. Rather, they see dominance or importance of the human user in these processes, which are still seen as “human-centered” (the human actor gives the decisive steering impulse) and/or “machine augmented” (the machine actor complements or augments the actions of the human actor in a complex, consistent, and context-related way). Depending on the intensity of the support and the actor that takes the assistance role, a distinction could be made between human-in-the-loop (a human actor being assisted by a machine) or machine-in-the-loop (a machine assisted by a human). In the interaction fields of GRAI both situations could arise.

The socialization interaction field in GRAI

Socialization is a dialogue-oriented setting with the primary intention of knowledge sharing between two actors, including the comprehension of other viewpoints and opinions. It is a highly contextualized process that is transmitted using natural language.

From the perspective of the machine agent toward the human agent, it can be seen as a setting that is oriented toward knowledge or information acquisition for a human user, e.g. explaining a topic to a human user.

The socialization interaction from the human agent to the machine agent is another dialog-oriented situation with the focus on specifying a complex information demand or situation, e.g. in an extensive and possibly iterative prompting interaction that informs about a situation (e.g. providing a richer context for the dialogue).

The externalization interaction field in GRAI

Within the externalization interaction field, the main focus is the connection between the internal knowledge representation models and the physical and digital reality, with the data and information coming from there. This transition process usually requires substantial efforts to integrate new information into existing models or to make information accessible for those models. This aspect refers to human users and IT systems (machine agents) that should deal with new information that is often contextualized and vague/inconsistent. Generative AI offers new capabilities to bridge this gap in a more efficient way (without the need to build specific IT solutions) and in a more effective way. That is, being able to work with vague/inconsistent information due to the contextualized information processing capabilities of large language models (LLMs).

More recent versions of LLM-based systems such as ChatGPT or Google Gemini allow the human user to add additional relevant materials into the conversation, e.g. using the “memory function” of those systems. This enlargement of the context relates to the interaction field human agent to machine agent as it helps the machine to identify a more precise context with a domain specific focus in the dialogue. This way a general conversation can be leveraged toward a specific direction by the human user through providing externalized information to the machine.

Another application use case that is emerging rather frequently here is the use of generative AI in retrieval-oriented tasks using retrieval augmented generation (RAG), which combines an initial search query with specific document collections (externalized information) to derive more specific search results.

Turning to the use-case of information retrieval in enterprise specific information sources (enterprise search) in the field of KM, the use of generative AI could introduce a bias in the results originating from the foundation models used. However, these effects might be similar to the contextual integration of a human user and could be counteracted by carefully selecting and adopting the right foundation model.

Another interesting interaction field is machine agent toward the human agent where the generation of structured content comes from existing unstructured information that contains this information only in an implicit form. Generative AI with multimodal capabilities can use recognition functions for objects and their attributes and generate product information in a table-like structure. The results could then be used for product catalogs, e.g. in the e-commerce domain.

The internalization interaction field in GRAI

Internalization relates strongly to the creation of internal models (or representations) of the knowledge of the outer world of the agent, reflecting the views and beliefs of the agent. The assistance of generative AI might both help the internalization processes of the human agent and also be beneficial to proactively build or adapt (digital) representations for the machine agent.

The direction from the machine agent toward the human agent is the process of supporting the internalization of a human user with the help of generative AI, e.g. creating a better understanding of a certain concept or topic.

The opposite direction human agent to machine agent is a situation in which the generative AI has an observing role for a longer period of time to suggest appropriate support actions for the human user based on the internal model built from those observations. Such situations could be a moderating role during online meetings (including summarization and analytics of the discourse) or the general user support across different applications, e.g. the Copilot functionality in Microsoft Office 365 products.

The combination interaction field in GRAI

The interaction field of combination might have received the most attention with the advent of generative AI systems because the way that LLMs could generate content that was combined from a large source of data was unique with the advent of ChatGPT. It did not require specific expertise to access this functionality – requests for the combination of externalized information could be stated in natural language with all its ambiguity.

With respect to the interaction field machine agent toward the human agent, the generation of complex summaries on a subject or provided information sources was one of the most prominent examples. The combination task could be configured for a specific information demand, given style, or tone of text (e.g. content generation for a specific target group) to be used by the human agent for further processing. Another example in this interaction field could be the generation of meeting protocols from the transcripts of online meetings.

Likewise, the flexibility of combining content elements in very different ways could also be used as a creativity tool for the human user (human agent to machine agent). Here, the human user combines different subjects in a single prompting request or dialogue sequence and therefore requires the generative AI system to combine different patterns that are otherwise unlikely to appear in the existing reality, e.g. the generation of images in the style of a certain artist, or imagery that combines aspects that are not existent in the physical reality.

2. Hybrid SECI framework (Cerchione, Liccardo, & Passaro)

Roberto Cerchione, Giuseppe Liccardo, and Renato Passaro propose their hybrid SECI framework in a 2026 open access Creative Commons paper¹³ published in Journal of Innovation & Knowledge. The framework is another of the outputs of a bibliometric literature review analysing 1411 articles related to the topic of Gen AI in KM, with this same research having also brought us the very valuable GenAI-KM organizational readiness self-assessment tool¹⁴ with further insights yet to come.

Cerchione, Liccardo, & Passaro contend that the disruptive potential of GenAI challenges the SECI model at multiple stages. For instance, GenAI facilitates the automatic externalization of fragmented ideas into coherent explicit formats, a task traditionally performed by humans. Similarly, GenAI has the capacity to combine vast volumes of structured and unstructured data in ways that mirror, or even exceed, human combination processes. It can also assist humans in the internalization process by providing explanations and tailored content that facilitate users’ experiential understanding. In this way, GenAI acts as a supportive environment rather than as a participant, simulating a form of machine-guided learning even without real tacit absorption.

Moreover, GenAI tools, particularly conversational agents, may create the illusion of socialization, mimicking interpersonal knowledge exchange through dialogue-like interactions. However, these interactions suffer from limited epistemic depth, as they are not grounded in shared human experience. While this surrogate of socialization has some positive aspects, it also highlights the need to distinguish between the effectiveness of simulated dialogue and genuine tacit-to-tacit knowledge transfer.

Table 1 summarises the impact that Cerchione, Liccardo, & Passaro find that GenAI has on the SECI phases. They advise that their observations point to a hybrid extension of the SECI model, where human and artificial actors co-create, validate and iterate knowledge across overlapping yet distinct learning loops.

Table 1. GenAI impact on SECI phases. Source: Cerchione, Liccardo, & Passaro, 2026.

Cerchione, Liccardo, & Passaro find that, at a deeper level of analysis, many contributions in the literature implicitly point to the consideration that GenAI is not only automating aspects of KM but also reshaping its foundational logic. Treating GenAI purely as a support technology appears too reductive. Instead, the evidence suggests that this disruptive technology introduces a new epistemic dimension to KM, in which knowledge can be generated through human-machine interaction. They further explore this consideration by proposing the existence of a machine dimension in knowledge generation, operating in a parallel manner with the human dimension described in the SECI model.

The machine dimension

Cerchione, Liccardo, & Passaro state that the emergence of GenAI as a non-human actor in knowledge processes calls for reconsideration of foundational KM frameworks. Results from their study suggest that GenAI does not merely support traditional KM processes, it introduces a machine dimension of knowledge creation that operates parallel to the human one. This conceptual expansion becomes especially evident when visualizing how GenAI maps onto, or diverges from, the four traditional SECI phases (as illustrated in Figure 3).

In specific SECI phases, GenAI aligns with existing knowledge creation processes. This is the case when it assists users in externalizing tacit ideas into structured outputs (Figure 3, number 2) and facilitates internalization through scenario simulation (Figure 3, number 4). However, other impacts diverge from traditional SECI logic, revealing knowledge dynamics better situated within a machine-driven epistemic loop. For instance, there is conversational prompting, which simulates aspects of socialization (Figure 3, number 1). Similarly, GenAI’s ability to produce synthetic datasets and synthesize patterns across them (Figure 3, number 3) reflects a form of synthetic combination that operates independently of human cognitive structuring.

These divergences suggest that GenAI is not simply participating in human knowledge creation cycles but is generating novel knowledge dynamics through a separate machine-driven loop.

This hybrid SECI framework provides a foundation for future knowledge management models that reflect the interplay between human cognition and artificial intelligence. More specifically, the machine dimension operates in a distinct but interconnected layer with respect to the human spiral. As schematized in Figure 4, the machine dimension can be framed as a complementary loop, distinct from, but potentially interlinked with, the human SECI process.

3. GenAI SECI model (Uchihira, with Nishimura & Ijuin)

Naoshi Uchihira proposes the GenAI SECI model in a 2026 open access Creative Commons paper¹⁵ published on the arXiv preprint platform. The model was developed from a literature review, including previous research by Uchihira and colleagues Takuichi Nishimura and Koki Ijuin, and observations of workplace knowledge transfer in Japan.

Uchihira first clarifies the knowledge to be managed – specifically, workplace knowledge, or “Gen-Ba” knowledge¹⁶ (Uchihira et al., 2023)). In Japanese, “Gen” means “actual and physical field,” and “Ba” means “knowledge creating space,” which was introduced by Nonaka & Takeuchi. Gen-Ba includes various settings such as manufacturing sites, maintenance and inspection sites, medical and nursing care settings, agricultural sites, and sales and customer service environments. Furthermore, knowledge here is defined as justified belief that serves as the basis for human action. Here, “justified” means “something that one has genuinely accepted and internalized for oneself.”

Uchihira classifies Gen-Ba knowledge into three layers: explicit knowledge, latent knowledge, and tacit knowledge (Figure 5).

The three layers of knowledge are continuous and do not have clear boundaries between them:

• Explicit knowledge: Knowledge that is consciously recognized and can be codified (manualized) in a complete and systematic manner.
• Latent knowledge: Knowledge that is not ordinarily conscious, but can be partially and fragmentarily expressed as code (text, image, video, etc.) when one is in the Gen-Ba or when asked by others.
• Tacit knowledge (narrow sense): Knowledge that exists unconsciously and cannot in principle be expressed in code (verbalized). A representative example of tacit knowledge is embodied knowledge such as the skilled techniques of craftspeople.

The tacit knowledge of the SECI model encompasses both latent knowledge and tacit knowledge (narrow sense), and is referred to by Uchihira as tacit knowledge (broad sense). Uchihira references Harry Collins¹⁷ who classified tacit knowledge into “relational tacit knowledge”, “somatic tacit knowledge”, and “collective tacit knowledge”. Somatic tacit knowledge corresponds broadly to tacit knowledge (narrow sense), while relational tacit knowledge corresponds to latent knowledge. Regarding collective tacit knowledge, those elements that can be fragmentarily codified are included in latent knowledge, while the remainder are classified as tacit knowledge (narrow sense).

It is Uchihira, Nishimura, & Ijuin’s view that much of collective tacit knowledge can in fact be partially codified. With respect to somatic tacit knowledge as well, those elements that can be fragmentarily codified may reasonably be included in latent knowledge. Here, codification is defined to include meaningful text, images, video, and similar forms of expression.

Furthermore, the proposed GenAI SECI model introduces “digital fragmented knowledge”. Explicit knowledge, latent knowledge, and tacit knowledge (narrow sense) are classifications of knowledge from a human perspective. Digital fragmented knowledge is defined as knowledge in cyberspace – specifically, explicit knowledge and latent knowledge that has been accumulated in cyberspace in some digital form. For example, records of what people in the Gen-Ba have sensed or thought, documented in language or photographs, do not constitute systematic knowledge like a manual, but can be accumulated in cyberspace as partial and fragmentary knowledge (referred to as “knowledge fragments”). While such partial and fragmentary knowledge has traditionally been difficult to handle within knowledge management, generative AI has now made it possible to utilize them. Figure 6 shows the proposed GenAI SECI model.

In this GenAI SECI model, each process is defined as follows:

• Socialization: Same as socialization in the SECI model. The process by which Gen-Ba knowledge is shared within an organization, through direct experience sharing, observation, and imitation.
• Externalization: The process of codifying Gen-Ba knowledge – even partially and fragmentarily – as “knowledge fragments” and accumulating them in cyberspace.
• Combination: The process of organizing the “knowledge fragments” accumulated in cyberspace into a form suitable for internalization.
• Internalization: The process of referencing the “knowledge fragments” accumulated and organized in cyberspace to amplify human Gen-Ba knowledge.

The roles of generative AI in externalization, combination, and internalization of the GenAI SECI model are as follows:

• Externalization: When codifying Gen-Ba knowledge distributed across various media (text, sensor data, images, video, etc.) as “knowledge fragments,” generative AI aggregates them into meaningful units.
• Combination: Generative AI organizes and stores “knowledge fragments” in a structured format such as a knowledge graph.
• Internalization: Generative AI selects and presents “knowledge fragments” that are effective for amplifying human Gen-Ba knowledge.

Specific technologies and functions are provided in the work-in-progress digital knowledge twin system¹⁸ (Figure 7).

The distinctive feature of the GenAI SECI model is that, rather than externalizing tacit knowledge into complete explicit knowledge and then using that explicit knowledge for internalization as in the SECI model, it leverages generative AI to amplify Gen-Ba knowledge through internalization using incomplete knowledge fragments, without requiring full externalization into explicit knowledge.

Header image: GenAI SECI model. Source: Uchihira, 2026.

References: