Chapter 5 - Political Economy of Generic Models: Sovereignty and Extraction
5.1 AI mediation and the political economy of knowledge
This chapter develops the political-economic component of the thesis. Where Chapter 4 described how expertise is materialised in European infrastructure advisory through the integrated three-scale system, this chapter examines what happens to that integrated system under the political-economic conditions of contemporary AI mediation. Two parallel arguments run through the chapter: a sovereignty argument about identity at risk under generic mediation operating at two scales, and an extraction argument about the dark mirror of the SECI model when firm knowledge flows outward into vendor systems.
The two arguments are connected. The same Stieglerian mechanism - dis-individuation under generic technical mediation - operates across multiple scales, and the structural risks the chapter analyses are different manifestations of one underlying dynamic. Generic AI mediation does not respect the integrated three-scale system; it operates across the scales, tilting individual practitioners toward common defaults, eroding firm-distinctive ways of seeing, and homogenising the texture of advisory work and eroding what is distinctive about European public-interest practice. Annex D develops individuation as the philosophical concept through which the framework engages the multi-scale mechanism analytically.
The chapter is the most politically charged in the thesis. It engages claims about the cultural and ideological character of major AI labs, about the structure of the model-vendor industry, and about the conditions under which firm knowledge is exposed to extraction through AI-mediated workflows. These are claims where the empirical literature is uneven, where the most pointed sources are journalistic, not peer-reviewed, and where the framework I deploy is articulating analytical commitments that the existing literature on AI in consulting has not engaged. I treat the politically charged claims with care, distinguishing structural argument from documented empirical practice, hedging where the evidence does not support stronger formulations, and registering when claims are anecdotal, not load-bearing.
The chapter also includes the defence reference case as an institutional anchor. European policymakers already treat AI sovereignty as a strategic concern in defence, where the logic of sovereignty is institutionally recognised. The defence case provides a reference point for extending sovereignty thinking to other infrastructure subdomains - energy, water, transport - where the same logic applies but the institutional recognition is less developed. The reference case is offered not as a model to be copied but as evidence that the sovereignty argument the chapter develops is institutionally available within European policymaking.
5.2 Sovereignty as identity at risk: the national-cultural scale
The thesis’s sovereignty argument works across two scales, and the chapter develops both before showing how they are connected through a single Stieglerian mechanism. The national-cultural scale is the scale at which most of the existing literature on AI sovereignty operates, but the existing literature addresses predominantly the procedural level - data residency, cloud infrastructure, regulatory compliance - not the level I treat as central.
The sovereignty argument is this: generic large language models are produced within particular cultural and discursive contexts, and the patterns of cultural-discursive material that constitute their training distributions carry the cultural-discursive patterns of their production contexts into the texture of any work the models mediate. This is not a claim that any particular model encodes any particular ideology in any direct way. It is a claim about distributional weight: the registers, framings, exemplars, and tacit assumptions that dominate a model’s training distribution are the registers, framings, exemplars, and tacit assumptions that the model produces most fluently and that mediate-using practitioners absorb most readily.
The cultural-discursive patterns of contemporary general-purpose LLMs are predominantly Anglo-American. This is not a controversial empirical claim. It is documented in the literature on cultural bias in language models, in the literature on training data distributions, and in the literature on AI ethics where the dominance of English-language and Anglo-American cultural material is widely registered. What the framework adds is a reading of what this dominance implies for European infrastructure advisory specifically.
The European public-interest tradition in infrastructure has a distinctive character that Chapter 4 described: a tradition of public-purpose engineering, a regulatory architecture for utilities, an intensifying EU governance frame around climate transition and strategic autonomy, and a third-box exposure (developed in Annex B) that operates differently in Europe than in markets where infrastructure is more thoroughly commercialised. This tradition is not encoded in the dominant patterns of generic AI training distributions. The texture of European utility regulation, the political weight of public consultation processes, the institutional weight of climate commitments, the specific content of European strategic autonomy thinking - these are present in the training distributions, but they are not dominant patterns. They are minority patterns competing with the dominant Anglo-American patterns for the model’s expressive weight.
The consequence is structural. When European infrastructure advisory work is mediated by generic AI, the work is being mediated by a system that produces Anglo-American patterns most fluently. Practitioners who use the system absorb its dominant patterns more readily than its minority patterns. The texture of the work tilts, gradually and structurally, toward the dominant registers of the model’s training distribution and away from the registers distinctive to the European public-interest tradition. The tilt is not dramatic on any single engagement, but it is cumulative across engagements and across practitioners and across firms, and it is the form of what I treat as the national-cultural scale of sovereignty.
This is not a claim that the European tradition is being deliberately attacked or that the AI labs are pursuing a cultural-political project against European public-interest thinking. The claim is structural, not intentional: the dominance of Anglo-American patterns in training distributions produces dis-individuating pressure on minority traditions through the cumulative weight of distributional dominance. The mechanism operates regardless of intent. The political-economic conditions under which generic models are produced - predominantly in Anglo-American contexts, with predominantly Anglo-American training data, by predominantly Anglo-American firms - produce the distributional dominance, and the distributional dominance produces the dis-individuating pressure on minority traditions.
The thesis is careful here about claims that go beyond what the available evidence supports. There is a literature on the ideological character of major AI labs, particularly on the political views of certain prominent figures in the AI industry. The literature is partly journalistic and partly academic, and it engages questions about the influence of particular ideological currents - the libertarian-adjacent, longtermist, and effective-accelerationist strands sometimes associated with the Thiel network and broader Silicon Valley political economy. These ideological currents are real and are documented, but the claim that they directly determine the cultural-discursive patterns of model training distributions is harder to support than the structural claim about distributional dominance. The framework relies on the structural claim, which is well-supported, and treats the ideological-content claims as supplementary, not load-bearing. Where the chapter cites work on the ideological character of major AI labs, it does so to register that the political-economic conditions of model production are not neutral, not to advance the stronger claim that any particular ideological content is being directly transmitted through the models.
5.3 Sovereignty at the organisational scale: firms losing their distinctive voice
The second scale at which the sovereignty argument operates is the organisational scale, and this is the scale at which the framework’s contribution against the existing literature is most direct. The existing literature on AI in consulting addresses task and workflow levels; it does not engage organisational-scale sovereignty as a concern. The framework treats it as important because the firm’s distinctive way of seeing - the content that makes one consulting firm meaningfully different from another even when they sell similar services - is what the framework Chapter 4 described as organisational individuation.
Madrid, 2025. A Spanish infrastructure advisory firm has built its reputation on its specific reading of the regulatory dynamics between Spanish network operators and CNMC - a reading formed through two decades of post-liberalisation experience, through specific relationships with staff across the regulatory body’s directorates, through a body of engagement knowledge that the firm’s senior consultants carry as professional identity. An Applied-stage consultant on a gas transmission project produces a section on regulatory risk framing for a client report. It is structurally correct. It is well-written. It covers the right categories in the right order. It reads as if written by a McKinsey team advising a generic European national regulatory authority client. The firm’s managing partner reads it and asks whether it was AI-assisted. The consultant says: partly - he used an LLM to structure the regulatory landscape section and then edited for firm-specific framing. The managing partner does not say this is wrong. He says: the client is paying for our reading of CNMC, not a reading of what a generic NRA would do. Those are different things. I need to see the difference in the document.
The mechanism at the organisational scale is structurally analogous to the national-cultural mechanism but operates within the consulting profession rather than across cultural traditions. Generic AI tools used uniformly across firms produce outputs that converge toward common defaults - the standardised analytical structures, the standard framings, the standard vocabulary that the model has absorbed from the wider corpus of consulting work. The firm’s distinctive way of seeing - what Chapter 4 described as the content reproduced through the SECI cycle, embodied in firm-specific framings, methodological reflexes, sectoral instincts, and the particular vocabulary in which the firm recognises a situation - is structurally at risk under generic mediation that produces standardised outputs as a default.
The structural risk is not that the firm’s distinctive way of seeing disappears overnight. It is that the gradient of the work tilts toward common defaults under cumulative AI-mediated production, and what makes the firm distinguishable becomes harder to demonstrate through the artefacts that have historically demonstrated it. The Foundation-stage consultant who produces an output through generic AI mediation produces an output that reads as competent management consulting work - but it reads as competent management consulting work, not as competent work in the firm’s distinctive style. The Applied-stage consultant who reviews the output is reviewing through their own AI-mediated review, often using the same generic tools that produced the output. The Chartered-stage consultant who eventually signs off is operating on a deliverable that has been produced and reviewed through processes increasingly mediated by tools whose default outputs are common across firms.
The firm’s distinctive way of seeing is not simply external to this process. It is reproduced through the processes that the AI mediation is increasingly absorbing - through the formative engagements at Foundation stage, through the supervisory engagements at Applied stage, through the senior-tier judgment at Chartered stage. When those processes are increasingly mediated by generic tools, the firm-distinctive content has fewer pathways through which to be transmitted, and the firm’s individuation under the SECI cycle becomes harder to sustain.
Benjamin’s concept of the aura - the quality of authentic presence that derives from a work’s embeddedness in a specific tradition, its “here and now” - provides vocabulary for what is at stake at the organisational scale. The AI-generated advisory output is a mechanically reproduced piece of advisory work in Benjamin’s sense: detached from the tradition it nominally represents, bearing the generic marks of the training corpus rather than the specific marks of the European public-interest tradition the work answers to. The firm’s distinctive way of seeing is an auratic property: it derives from the firm’s accumulated history, the sedimented judgments of its practitioners, the conventions of its analytical approach built across decades of sustained engagement with specific European regulatory institutions. What makes the firm’s output distinctively its own - recognisable to a client across multiple engagements, legible to a regulator who has negotiated with its practitioners - is not a brand or a style but the auratic quality of embeddedness: the traces of a specific tradition expressed through a specific firm’s sustained encounter with it. A system trained on the full range of consulting outputs cannot reproduce this aura; the generic training corpus averages across firm distinctions rather than deepening any one of them [@benjamin1935artwork]. The democratisation dimension should be held explicitly in view. Mechanical reproduction makes competent advisory output available at scale, and this is a real gain for analytical access across the profession. What is democratised is competence - adequate output, information in Benjamin’s sense, technically sound and analytically tractable. What cannot be mechanically reproduced is the auratic quality through which counsel derives its authority from the adviser’s formation within a specific tradition. The gain and the loss are real and simultaneous; the thesis’s normative argument is about what the gain cannot substitute for.
This is the rigorous form of what the informal discourse calls the “digital twin” concern. The informal version of the concern is that AI tools will learn how a particular practitioner works and replace them. The rigorous version is structural: the firm’s organisational individuation, the substance of what makes the firm valuable across generations of practitioners, is increasingly mediated through tools whose default outputs are common across firms, and the cumulative weight of generic mediation produces dis-individuation at the organisational scale just as it produces dis-individuation at the national-cultural scale.
The two scales are connected by the same mechanism. Generic AI tools, because they are generic, do not respect either national-cultural or firm-specific distinctiveness. They produce outputs that gravitate toward distributional averages, and the cumulative effect of working through them is dis-individuation at multiple scales. The framework’s analytical contribution is partly to make this multi-scale character visible. The existing literature on consulting under AI mediation can see efficiency gains and skill retention; it cannot see what is happening to the firm’s distinctive voice across generations because it does not have the analytical apparatus to read organisational individuation as a concern.
The chapter is careful here about the empirical reality. Firm distinctiveness is not yet visibly collapsing into common defaults. Major consulting firms continue to compete on differentiation, and their senior practitioners continue to articulate distinctive analytical positions. The structural pressure I name is operating cumulatively through AI mediation, not yet fully manifest but important. Whether the pressure resolves into actual organisational dis-individuation depends on conditions I also treat as variable: the political-economic conditions of model production (which Chapter 5 §5.4 develops), the institutional choices firms make about how AI mediation is integrated into their workflows, and the wider trajectory of European infrastructure advisory which Chapter 8 examines.
5.4 Extraction: the dark mirror of the SECI model
The second argument the chapter develops is the extraction argument, and this is where the framework’s most pointed political-economic claims operate. The extraction argument extends the SECI model that Chapter 4 introduced to characterise organisational knowledge management in consulting firms - and reads it through what I have called the dark mirror, where the firm’s externalised knowledge no longer recirculates within the firm’s own SECI cycle but flows outward into vendor systems and back into the consulting market through generic capability.
London, 2024. An infrastructure advisory firm is deploying an LLM vendor’s document summarisation tool for regulatory filing review - genuinely useful, commercially defensible, and significantly faster than manual review. A senior partner asks the technology team to check the vendor’s standard enterprise agreement before the firm processes client materials through the tool. The IT procurement lead’s initial response is that the standard enterprise tier includes a training-data clause. The partner escalates. A three-week legal review follows. The firm ends up on a bespoke agreement that explicitly excludes client content from the vendor’s training distributions, with audit rights and breach notification obligations. The partner’s question cost the firm additional legal fees and a delayed deployment. Three months later, a peer firm in the same market receives a client request to confirm that the firm’s AI tools do not use client content for training. The peer firm does not have a clear answer. The partner’s firm does.
The historical assumption underlying the SECI model, and underlying organisational knowledge management more broadly, is that the explicit artefacts produced through externalisation remain within the firm. The methodology document is the firm’s intellectual property; the playbook is the firm’s competitive asset; the case study is the firm’s institutional memory. These artefacts are transmitted across the firm’s generations of practitioners through internalisation, and they are protected from competitors through institutional architecture: employment contracts, confidentiality agreements, internal-only access, and the practical difficulty of reproducing tacit firm context outside the firm.
Contemporary AI mediation introduces a structural pathway through which this assumption can fail. When firms adopt AI-mediated workflows that involve sending working materials - drafts, frameworks, case-relevant documents, internal analyses - through systems operated by external vendors, the firm’s externalised knowledge is no longer purely within the firm’s control. The vendor’s terms of service, data-handling practices, and use of customer content in training subsequent models all bear on whether the firm’s externalised knowledge stays within the firm or flows outward.
The empirical reality varies by vendor, by contract, and by deployment configuration. Some major vendors offer enterprise contracts that explicitly exclude customer data from training. Some firms use deployment configurations - on-premises models, vendor-hosted but isolated tenant environments, retrieval-augmented systems that keep firm content within firm boundaries - that mitigate the structural risk. Other vendors and other configurations expose customer content to training, often without the customer’s full understanding. The structural risk exists; the empirical realisation varies.
The framework’s claim is structural rather than evidentiary. I do not claim that all firms’ externalised knowledge is currently being extracted into vendor training distributions. I claim that the structural possibility now exists in a way it did not before, that the broader pattern of which extraction would be an instance is well-documented in adjacent domains - the surveillance capitalism literature, the data colonialism literature, the platform governance literature all examine variants of this structural dynamic - and that the consequences if extraction occurs in the consulting domain are significant enough that the question is worth treating as a strategic concern. The frameworks that the existing IS literature on consulting deploys do not engage extraction as a concern at all; the framework’s analytical contribution is partly to make it visible.
The dark mirror reading of the SECI model clarifies what is at stake. In the original SECI model, the externalised knowledge produced by senior practitioners - methodologies, playbooks, frameworks, case studies - feeds back into the firm through internalisation, where junior practitioners absorb the explicit artefacts into their tacit competence. The firm’s individuation is reproduced across generations through this cycle. In the dark mirror reading, the externalised knowledge flows outward into vendor systems whose training distributions absorb it, and the absorbed content is recirculated back into the consulting market through the generic capabilities of the vendor’s models - capabilities that are then sold to the firm itself, to the firm’s competitors, and to the firm’s clients in their development of in-house capability.
The structural concern is that the firm’s distinctive externalised knowledge becomes part of the generic distribution that competitors and clients then access. The content that has historically been the firm’s competitive asset - what the firm has individuated itself through - is, structurally, at risk of being absorbed into the common ground that AI mediation operates on. The threat is not to any individual consultant but to the firm as a knowledge-managing collective whose distinctiveness depends on its externalised knowledge remaining within its own SECI cycle.
I treat the most pointed claims in this section with care. The claim that data brokers systematically purchase Slack archives, email and chat logs from failed startups to feed into LLM training is one I have heard in informal professional discussion but cannot fully ground in peer-reviewed sources. The bankrupt-asset data sale market is documented, including in FTC interventions in specific cases, and the data-broker pipelines feeding model training are documented in the broader surveillance and platform-governance literature. But the specific claim about consulting firm working materials moving through these channels is anecdotal in the empirical record I have access to. I treat it as illustrative of the structural risk demonstrated by analogous cases in adjacent domains, not as documented current practice in consulting specifically. The framework’s argument depends on the structural pattern being real and the consequences being significant, both of which are well-supported; it does not depend on every anecdotal claim about specific extraction pathways being verifiable.
The Thiel/Silicon Valley political economy framing is also engaged with care. The political views of certain prominent figures in the AI industry are documented, including views that the framework has concerns about - libertarian-adjacent, longtermist, effective-accelerationist, and in some cases overtly anti-democratic political positions. These views are present in the political-economic context within which generic AI is produced. Whether they directly determine the cultural-discursive patterns of model outputs is harder to establish than the structural claim about distributional dominance. The chapter cites work on the political economy of AI labs to register that the production context is not neutral, but it relies on the structural claim about distributional dominance for the sovereignty argument rather than on stronger ideological-content claims that the empirical literature does not yet support.
5.5 The defence reference case
The defence reference case provides an institutional anchor for the chapter’s sovereignty argument. European policymakers already treat AI sovereignty as a strategic concern in defence, where the logic of sovereignty has been institutionally recognised for decades. The defence case is offered as a reference point - not as a model to be copied directly, but as evidence that the sovereignty thinking the chapter develops is institutionally available within European policymaking, and that the conceptual move from procedural to sovereignty has already been made in at least one infrastructure subdomain.
Warsaw, 2023. An advisory team is conducting a rail network resilience assessment for PKP - standard infrastructure methodology: redundancy analysis, demand scenarios, maintenance investment prioritisation. In the third client meeting, a figure whose position is listed on the briefing note as “infrastructure coordination” from the Ministry asks a set of questions the team has not prepared for: not about passenger demand or maintenance cycles, but about freight corridor capacity under disruption scenarios, alternative routings for bulk materials, and recovery timelines for specific segments of the east-west corridor. The team lead recognises, without anyone saying so, that these are not transport planning questions. The engagement subsequently re-scopes. The resilience methodology the team has developed - designed for a commercial transport planning context - is not the methodology that applies when the infrastructure is also a strategic military logistics asset. The team leader has never been told this explicitly. She understands it.
The sovereignty argument in defence has historically been about strategic autonomy - about the conditions under which European defence capability is dependent on systems controlled by extra-European entities, and about what dependency means for the autonomy of European defence decisions. The argument has been instantiated through institutional architectures: domestic procurement preferences, technology transfer requirements, restrictions on foreign ownership of strategic defence firms, and increasingly through specific provisions on AI capability in defence contexts. The logic is that defence capability cannot be subject to the operational decisions of entities outside the European political community without compromising the sovereignty of European defence policy.
The framework’s claim is that the same logic applies to other infrastructure subdomains where the European public-interest tradition is constitutive of the domain’s content. Energy infrastructure decisions are not strategically equivalent to defence decisions, but they are nonetheless decisions whose content depends on European public-interest commitments - to climate transition, to social tariffs, to regulated affordability, to consumer protection - that the sovereignty logic engages. Water infrastructure, transport, urban planning, and the other subdomains in the family Chapter 4 described all share, to varying degrees, this characteristic.
The institutional recognition of sovereignty in these other subdomains is uneven. Some EU policy has begun to engage AI sovereignty in non-defence contexts - the European AI Act, certain provisions of the Green Deal, some procurement frameworks - but the recognition of sovereignty as constitutive of European public-interest practice in infrastructure is less developed than in defence. The framework’s contribution is partly to articulate the logic that would extend defence-style sovereignty thinking to other infrastructure subdomains, and Chapter 9 develops the recommendations to European policymakers that this articulation supports.
The chapter is careful here about overreach. The framework does not claim that all infrastructure subdomains require defence-equivalent sovereignty architectures. Defence has features - secrecy requirements, kinetic capability, military application - that other subdomains do not share, and the institutional architecture of defence sovereignty cannot be transferred wholesale. My claim is that the logic of sovereignty - the recognition that the political community has standing in decisions that affect its autonomy - is institutionally available in defence and applicable, in modulated form, to other infrastructure subdomains. The modulation is part of what Chapter 9 develops in the policymaker recommendations.
5.6 The integrated political economy
The two arguments the chapter has developed - sovereignty and extraction - are connected through a single underlying mechanism: dis-individuation under generic technical mediation, operating at multiple scales of the integrated three-scale system.
The mechanism is consistent across the chapter’s arguments. At the national-cultural scale, generic AI mediation tilts the texture of European infrastructure advisory toward dominant Anglo-American patterns, eroding the distinctiveness of the European public-interest tradition. At the organisational scale, generic AI mediation tilts firm-level outputs toward common defaults, eroding the firm’s distinctive way of seeing as reproduced through the SECI cycle. At the level of the firm’s externalised knowledge, generic AI mediation creates pathways through which firm-distinctive content flows outward into vendor systems whose generic capabilities are then recirculated to competitors and clients. The mechanism is one mechanism, operating at multiple scales, producing dis-individuating pressure on the content the integrated three-scale system Chapter 4 described.
A convergence with formal economic theory is worth registering at this point, because the political-economic argument I have developed across the chapter is not the only register in which it can be made. The economist Daron Acemoglu and his co-authors Dingwen Kong and Asuman Ozdaglar [@acemoglu2026collapse] develop a formal dynamic model of human learning under agentic AI that arrives at conclusions structurally aligned with the framework I have been deploying. Their distinction between general knowledge (community-level, accumulated, complementary to human effort) and context-specific knowledge (individual, idiosyncratic, substitutable by agentic AI) maps onto a distinction the thesis draws between expertise reproduced through the integrated three-scale system and interactional fluency produced by AI mediation. Their learning externality - that human effort jointly produces both private decision-quality and a public good (the community’s general knowledge stock), and that the public good is what makes future human effort productive - gives formal economic backing to what the thesis argues philosophically: that the integrated system through which European infrastructure advisory has historically reproduced its expertise depends on the formative engagements through which that expertise is transmitted, and that AI mediation as substitute for those engagements depletes the ground on which future advisory work would rest. Their central result, that under sufficiently accurate agentic AI the system can tip into a knowledge-collapse steady state where the community’s general knowledge vanishes despite high-quality individual decisions, is the formal-theoretic analogue of the political-economic argument developed above. The two arguments operate in different methodological registers - theirs in formal dynamic equilibrium analysis, the thesis’s in philosophical-political-economic analysis - but they converge on the same structural concern.
The convergence is informative but the methodological registers differ in ways the thesis treats as substantive. Acemoglu et al. work at the level of information aggregation: their general knowledge is community-aggregated information, statistical in character, and the collapse is the depletion of an information stock. The thesis works at the level of formation and individuation: the expertise at risk is not statistical information but the contributory expertise reproduced through the SECI cycle, embodied in firm-distinctive ways of seeing, sustained across generations through the formative engagements that make junior practitioners into senior ones. The two registers are complementary, not equivalent. Acemoglu et al. show formally that an information-aggregation system tips toward collapse under specific conditions; the thesis shows philosophically that a system of professional formation faces structurally analogous conditions, and that what is at stake includes the European public-interest tradition rather than merely the aggregation of information about it. The formal economic argument supports the philosophical political-economic argument; it does not replace it.
The political-economic conditions that produce the mechanism are also consistent. Generic AI is predominantly produced by a small number of very large firms, predominantly headquartered in the United States, with predominantly Anglo-American cultural and discursive contexts of production, with training distributions dominated by the same cultural and discursive patterns, and with commercial incentives that favour the broad deployment of generic capabilities across markets. These conditions are the conditions under which the mechanism operates, and they are not given. They are political-economic conditions that could be different, and the framework’s political-economic argument is partly to make visible what would have to change for the conditions to be different.
The chapter has been careful to register where the available evidence supports the framework’s claims and where the claims are structural rather than evidentiary. The structural sovereignty claim - that distributional dominance produces dis-individuating pressure on minority traditions - is well-supported. The structural extraction claim - that the political-economic conditions of contemporary AI vendor relationships create pathways through which firm knowledge can flow outward - is well-supported. The empirical instantiation of these structural claims varies by jurisdiction, by vendor, by firm, and by deployment configuration; the chapter has registered this variation honestly while maintaining the structural claim I treat as central.
What the chapter has not done - and what subsequent chapters take up - is examine what the political-economic conditions imply for the stage-by-stage analysis of European infrastructure advisory (Chapter 7) and for the futures the profession faces (Chapter 8). The recommendations to the four constituencies are developed in Chapter 9 on the basis of the political-economic conditions the chapter has named, the European public-interest tradition Chapter 4 described, and the cross-scenario findings Chapter 8 develops.
5.7 The political-economic conditions
Two arguments establish the political-economic conditions under which the proletarianisation mechanism operates in European infrastructure advisory. The sovereignty argument - working across the national-cultural and organisational scales through the Stieglerian mechanism of dis-individuation under generic technical mediation - names what is at risk when advisory work is mediated by systems whose training distributions are dominated by Anglo-American cultural and discursive patterns. The extraction argument - the SECI cycle running outward rather than inward, with firm-externalised knowledge flowing into vendor systems and recirculated as generic capability - names the mechanism through which that risk is realised at the organisational scale. The defence reference case anchors both arguments in an institutional domain where European policymakers have already recognised what is at stake: the logic that applies in defence applies wherever AI sovereignty matters for European public-interest infrastructure.
These conditions are not given. Generic AI mediation operates in predominantly Anglo-American production contexts, with vendor structures that create extraction pathways and commercial incentives that favour broad deployment over European contextual specificity. That these conditions obtain at the moment of the thesis’s writing is a structural claim the framework makes; it is also a claim about conditions that could be different, and the sovereignty argument is partly to make visible what would have to change.
Chapter 6 asks whether the proletarianisation mechanism is visible in adjacent professions - software and law - where AI mediation has been operating at scale for longer than in management consulting, and what comparative evidence from those cases the framework can carry forward.