The Second Operating System of Power: Who Owns the Operating System – Part 2

by Michael Hollister
Published at apolut media on September 11, 2026

5.018 words * 27 minutes readingtime

Part 1 read here:
How Palantir Connects and Divides Germany’s Security Agencies – Part 1

The First Operating System of Power read here:
BlackRock, Aladdin, and the Invisible Infrastructure

The Business, the Lock-In, and the Question of Who Controls the Switch

Of all institutions, it is the state that provides the most reliable funding to the very corporation whose software calls much-vaunted digital sovereignty into question. In the second quarter of 2026 alone, U.S. government agencies paid Palantir $809 million – in just three months, and 90 percent more than a year earlier. Total revenue rose by 93 percent to $1.935 billion in the same quarter, net income climbed from $327 million to $1.06 billion, and the company raised its annual forecast to more than $8 billion. Government business is the oldest part of this company, the anchor around which everything else has grown. It is also the part that makes a simple truth visible: Anyone who procures an operating system for their most sensitive tasks does not pay once, but permanently.

The market has turned that permanence into a bet of dizzying proportions. In early September 2026, the stock was trading at around $183, with market capitalization between $430 billion and $440 billion – roughly 54 times expected revenue for the current year and a price-to-earnings ratio above 140. This valuation rests on a workforce of around 4,400 employees. Investor Michael Burry, known for betting against the U.S. mortgage market in 2008, renewed his short position against Palantir in early September and called it a bubble; just days later, an expanded alliance with consulting firm PwC pushed the share price up by 7 percent. This article does not decide whether the valuation is excessive or justified. It asks a different question.

Behind the numbers lies a structure. The first part of this analysis mapped what Palantir’s software does in Germany, who uses it, and who rejects it on fundamental grounds – and, through the U.S. CLOUD Act, opened a question that this second part addresses. That question is: Who profits from this operating system, who owns it, and what does it mean when a state delegates control over part of its data infrastructure to a publicly traded company whose value depends precisely on nobody ever being able to leave? The answer leads through the business model, a striking parallel in the financial markets, the limits of parliamentary oversight, and finally the question of who actually benefits from the so-called European path.

Invest, Then Harvest – How Support Becomes Dependency

Palantir was founded in the aftermath of the September 11, 2001, attacks, based on the diagnosis that the intelligence services had failed less because they lacked data than because they were unable to combine information scattered across fragmented systems. That integration remains the product to this day – and it cannot be sold off the shelf. Palantir does not deliver its software as a finished standard product that an agency installs and then operates on its own. The company sends engineers into the institution to adapt the platform to its structures, import data from its source systems, and keep the environment running. The industry calls these employees “Forward Deployed Engineers,” a term Palantir borrowed from the military language of forward-deployed forces. The first part of this series showed that this operating model was precisely why the Bundeswehr and the Federal Office for the Protection of the Constitution rejected the system: Whoever maintains the operating system potentially gains insight into what runs on it. From a business perspective, that same model is the core of the enterprise.

The economics follow a pattern analysts describe as an investment-and-harvest cycle. At the beginning of a customer relationship, Palantir deploys engineers at close to cost; margins are low and the effort is high. Once the platform has been tailored to the agency’s workflows, however, support requirements decline and the project business becomes a high-margin licensing business. What initially appears in aggregate financial statements as an expensive bespoke approach is in fact the construction of a bond that pays off later. Existing-customer retention is correspondingly high; in the second quarter of 2026, spending by existing customers rose by 57 percent year over year, and the company signed 220 contracts worth at least $1 million in that quarter alone, including 73 worth more than $10 million. What begins as a costly support relationship ends as a stream of recurring licensing revenue – an investment-and-harvest model reinforced by security clearances as an additional switching barrier.

The crucial point is what creates the lock-in. It does not arise primarily from contractual clauses, but from operational dependency. The platform grows into the institution’s operating structure. It binds trained employees who have learned to work within its processes. It generates data histories and interfaces with existing source systems that are calibrated to its formats. It shapes daily workflows from analysis to reporting. Where many customers require the same improvement, that improvement migrates into the shared platform – the gravel track of a bespoke solution becomes a paved road that accelerates every subsequent project. That paved road is the ontology, the digital model of people, places, and events described in Part 1. The more deeply an agency builds its work around it, the more expensive and risky departure becomes.

These switching costs are even higher for government customers. Technical barriers are compounded by security clearances, classification levels, and the political capital invested by the program managers who once pushed the deployment through. Baden-Württemberg’s five-year contract, worth around €25 million and documented in Part 1, contains no exit clause – and the state was already paying the provider before the software was even operational. Migrating data away from the platform is possible, but it is a major undertaking: It requires replacing a system that has by then come to determine how the agency works.

Intellectual honesty requires acknowledging the dissenting views on this model, because they come from within the industry itself. Charlie Gottdiener, CEO of software provider Anaplan, regards the embedded-engineer model as an effective sales tool that produces quick results but a poor long-term strategy that creates customer dependency and limited functionality; he expects it to decline. Other observers defend the approach: High switching costs, they argue, are not a deliberately constructed trap but the natural consequence of software intervening deeply in real-world processes, and a properly structured contract can limit them – if the agency itself owns the code, test datasets, and operating manuals and uses open standards. It is also noteworthy that the model is now being copied: In 2026, Amazon invested $1 billion in embedded engineering teams, and other artificial-intelligence providers are adopting it as well. Lock-in is therefore not unique to Palantir, but the logic of an entire class of infrastructure providers – which makes it not smaller, but larger.

Two Operating Systems, One Pattern

This logic is not new. An earlier analysis in this series dissected it elsewhere: in Aladdin, the technology platform operated by asset manager BlackRock that has become the operating system of the financial markets. BlackRock manages around $14 trillion in assets, more than any other company in the world; the technology division that includes Aladdin generated around $2 billion in revenue in 2025, one-quarter more than the year before. Yet its true reach lies outside the corporation: More than 1,000 organizations worldwide – pension funds, insurers, central banks, and competing asset managers – run their investment processes on the same platform, with a three-year retention rate of 98 percent. Palantir is the second of these private operating systems of power, and comparing the two sharpens the view of both – precisely because their differences are as substantial as their similarities.

The similarities lie in their architecture. Both systems bring together what was previously separate: Palantir links police case, communications, and registry data; Aladdin links portfolios, liabilities, and derivatives. Both place a model of reality over the raw data – Palantir’s ontology of people, places, and events, and Aladdin’s risk architecture, which BlackRock markets as a “whole portfolio” system. Both generate lock-in that is operational rather than contractual: An Aladdin implementation takes years, binds trained staff, and creates data histories that cannot easily be migrated. When Swiss banking giant UBS terminated an Aladdin contract inherited from Credit Suisse in July 2025, it had to migrate hundreds of funds to its own systems – the rare proof that leaving is possible, but difficult. And both are black boxes: Neither Palantir’s algorithms nor Aladdin’s risk models are open to their users.

Most strikingly, all three describe themselves in related terms. In the English version of its ruling, the German Federal Constitutional Court described Palantir’s Gotham as an “operating system”. BlackRock calls Aladdin the “common language of the whole portfolio.” And the European alternative discussed later in this article markets itself as a “heterogeneous data operating system.” Three private operating systems, one term – and in every case, the description of a layer that does not manage the content itself, but the environment in which that content can be used in the first place.

The differences are equally important and must not be blurred. Aladdin sits in the financial markets; Palantir sits within the state’s coercive apparatus – police, military, and intelligence services. One system processes money; the other processes suspicion. The dangers identified by critics therefore differ. With Aladdin, the risk is systemic and financial: If hundreds of institutions calculate their risks using the same platform, they may reach similar conclusions, share the same blind spots, and dispose of the same positions simultaneously in a crisis. The British financial regulator warned that the failure of such a system could damage market integrity; a California pension fund rejected Aladdin’s risk functions, citing the danger of groupthink; and Klaus Holse, CEO of competitor SimCorp, reported that some customers avoided Aladdin because they did not want to enter the systemic risk of depending on a single provider. With Palantir, by contrast, the danger concerns fundamental rights: the linking of millions of datasets that also capture witnesses, victims, and uninvolved third parties, as Part 1 showed. One risk affects portfolios; the other affects people.

Both corporations are subject to U.S. law, but that matters more in Palantir’s case because the data involved are among the state’s most sensitive, rather than securities holdings. The two companies also develop their structural power by different routes. BlackRock advises its own regulators: In 2008, it managed around $130 billion in toxic securities from Bear Stearns and AIG on behalf of the Federal Reserve; in 2020, it purchased corporate bonds for the Fed under a program worth up to $750 billion. Palantir develops its power as its software is rolled out across the U.S. government, extending to the consolidation of government datasets cited in Congress as an example of citizen profiles being merged from separate sources. Two routes to the same position: a private company the state can no longer bypass.

Oversight That Runs Into a Dead End in Parliament

When a private operating system grows this deeply into the state, Parliament ought to oversee what happens there. Three minor interpellations from 2026 show where that oversight reaches its limits – and those limits do not lie at the margins, but precisely where the answers matter.

The first inquiry, submitted by the AfD parliamentary group and answered on March 23, 2026, concerned the possible use of Palantir by the Federal Police and the Federal Criminal Police Office. The German government first rejected the premise: No decision in favor of any particular software had been made. As a matter of principle, it declined to comment on details of ongoing technical assessments and market soundings, invoking a “core area of executive autonomy” that was not open to parliamentary investigation. The questioners also wanted to know whether such software would have access to health data from electronic patient records, population registers and traffic data, travel data, financial transactions, and communications, and what legal basis would permit that access. The government referred in general terms to existing powers and legislation still under consideration. Another subsidiary answer is revealing: Asked about the software’s quantifiable benefits, the government stated that it had no independent studies addressing the question. The executive branch thus both refuses to disclose information about the process and admits that it possesses no independent evidence of the benefits.

The second inquiry, also submitted by the AfD parliamentary group and answered on April 23, 2026, concerned the use of Palantir’s Gotham by German intelligence services. It asked whether data on politically active individuals, their families, members of Parliament, journalists, or attorneys were being processed. The government did not answer the central questions – not publicly and, as it expressly stated, not even in classified form for the Bundestag’s secure records office. Even a minimal risk of disclosure, it said, could not be accepted. The sentence with which it framed that refusal is noteworthy: The fact that an answer was being withheld should be interpreted “neither as confirmation nor as denial” of the matter in question. At least one piece of information was provided: Palantir software plays no role in the joint police and domestic-intelligence centers GETZ and GTAZ. As for constitutional assessments of its use by the intelligence services, the government said it had none.

The third inquiry, submitted by The Left parliamentary group and answered on July 08, 2026, concerned the Federal Office for the Protection of the Constitution’s purchase of ArgonOS analysis software from the French company ChapsVision. Twenty-one subquestions addressed precisely the points that determine fundamental rights and oversight: what data the system processes; whether the path to its results is traceable and reproducible; which sources it accesses; what environment it runs in; where the data are stored; who maintains the software; whether the testing phase involved individuals outside the agency; whether the contract was concluded directly with ChapsVision or its German partner rola; whether a tendering procedure took place; and what the purchase cost. The answer was the same as before: reasons of state, no response – not even in classified form for the secure records office. Disclosure, the government argued, would allow conclusions to be drawn about the agency’s technical capabilities and intelligence potential. Asked whether members of the German government owned Palantir shares, the government replied that ministers were not required to disclose assets and that it therefore held no centralized records. More generally, it stated that it supported neither a particular software product nor its rejection, but decided case by case; digital sovereignty had to be ensured through control over data and transparency in every solution.

Three inquiries, three mechanisms of refusal: the core area of executive autonomy for ongoing processes, reasons of state extending even to withholding information from the secure records office, and a declared software neutrality that reveals no commitment. Together they shut off precisely the questions that determine control over the operating system: which data, who operates it, at what cost, and under what oversight. This is not a scandal with a clearly identifiable culprit – there are good reasons for keeping intelligence capabilities secret. It is a structure: At the most sensitive level, Parliament learns the least.

The European Path and a Question That Remains Open

The Office for the Protection of the Constitution’s decision to choose ArgonOS rather than Palantir is regarded as the European path – a recovery of sovereignty through selecting a provider not subject to U.S. law. But who is the company that embodies this path?

ChapsVision was founded in mid-2019 by French software entrepreneur Olivier Dellenbach. According to the company, it employs more than 1,000 people, has over 2,000 major customers, generated around €200 million in revenue in 2024, and, like Palantir, is expanding through a targeted series of international acquisitions. In 2021, Dellenbach created a division called CyberGov, which bundles the ArgonOS system with artificial intelligence for cybersecurity, homeland security, defense, and intelligence. In June 2026, the French government also selected ArgonOS as its sovereign data platform. The system itself promises to transform complex, fragmented data into actionable intelligence – the same task performed by Palantir’s Gotham and, as noted, described using the same concept of an operating system. The Office for the Protection of the Constitution chose ArgonOS instead of Palantir for a reason documented in Part 1: Sovereignty over sensitive data holdings should remain within Europe, reducing dependence on U.S. service providers. ChapsVision accordingly markets its system as a powerful alternative to non-European providers that keeps control of sensitive data in Europe – the same language of sovereignty that shapes the German debate.

At this point, one piece of the mosaic requires careful handling, because it could easily be burdened with more significance than the evidence supports. Before Dellenbach founded ChapsVision, he established eFront in 1999, a specialist in managing private-equity and venture-capital portfolios. Over 20 years, eFront became the global market leader in its segment, with more than 850 employees, customers in over 30 countries, and revenue exceeding €130 million. In April 2019, Dellenbach sold the company – for $1.3 billion according to the parliamentary group submitting the inquiry, and for more than $1.5 billion according to ChapsVision’s own account. The buyer was BlackRock, which integrated eFront into its Aladdin platform and thereby entered the world of alternative investments, as an earlier analysis in this series documented. In that same year, 2019, Friedrich Merz was chairman of the supervisory board of BlackRock Asset Management Deutschland AG, a position he held from 2016 until March 2020.

The Left’s sixth question targeted this constellation. It asked whether Merz, in his BlackRock role at the time, had been in contact with Olivier Dellenbach during the eFront acquisition – the man who now heads ChapsVision, whose system the Office for the Protection of the Constitution selected instead of Palantir. The German government declined to comment, arguing that the question concerned an earlier role held by the current chancellor and was not subject to parliamentary scrutiny.

What is documented here must be stated precisely: a sale, the individuals involved, a parliamentary question, and a non-answer. What is not documented is any contact, connection, or influence on the Office for the Protection of the Constitution’s decision, which was based on the sovereignty considerations described in Part 1. Seven years and two independent events separate the sale of eFront to BlackRock from the selection of ArgonOS by the Office for the Protection of the Constitution; connecting them through one individual would exceed the chain of evidence.

The value of this piece of the mosaic lies elsewhere. It lies in the fact that an elected parliamentary group asked a specific question and the government left it unanswered by referring to the chancellor’s “previous role” – which did not resolve the question, but merely left it standing. The piece is placed alongside the others, not above them. Capability and structure are findings; intention remains an open question that this article neither answers nor seeks to answer.

Sovereignty as a Business Model

The European path is not a pure escape from the underlying problem, because it reproduces that problem structurally. ChapsVision does not deliver its platform in Germany alone. On October 15, 2025, the company entered into a strategic partnership with rola Security Solutions GmbH of Oberhausen, Germany’s market leader in security IT, and with IT service provider SVA. The division of labor is clearly defined: ChapsVision provides the technological foundation and continues developing the platform; rola handles integration and customization for German agencies – tailored analytical blueprints, data adapters, and training. Word for word, this is the same operating model that became contentious with Palantir: a private provider, an embedded integration partner, and lock-in through customization.

Rola, too, carries a history that places the concept of sovereignty in a distinctive light. Founded in Oberhausen in 1983, the company specializes in information management, networked case processing, and data analysis for security agencies; its systems are used by police, intelligence services, and the military. Since 2014, rola has been a wholly owned subsidiary of T-Systems International, which belongs to partly state-owned Deutsche Telekom. T-Systems expressly acquired rola at the time to prevent a foreign sale and the loss of expertise and, according to the rationale then given, to preserve “technological sovereignty” – in the context of the PIAV police information and analysis network, for whose central system rola was the federal government’s preferred candidate. The specialist publication Police-IT therefore calls rola a “state-created monopolist,” a characterization that should be attributed rather than adopted; the sober facts are that rola is the market leader, a Telekom subsidiary, and the provider of case-management systems to the Federal Criminal Police Office and the Federal Police.

The rhyme is unmistakable. The sovereignty argument used to justify the selection of ArgonOS in 2025 is the same one used to justify building a domestic monopoly in 2014. Sovereignty also functions here as industrial and location policy: It changes who earns the money, not whether a private operating system does. The American provider is replaced by a French one, integrated by a German Telekom subsidiary – a different network of interests, but the same basic structure.

The remaining question is whether there is any path out of this logic. It would have a name: P20, the Police 20/20 program. Originating in the Saarbrücken Agenda of the Conference of Interior Ministers, it is a joint federal-state project designed to place Germany’s fragmented police IT on a unified information architecture, with a shared data house at its core connecting all 20 German police organizations and around 320,000 employees; the main steps are scheduled for completion by 2030. The Federal Ministry of the Interior describes P20 as the common framework for this modernization. Initial applications developed under the P20 umbrella are already in use, ranging from digital crime-scene documentation to online police stations – evidence that the state can build software of its own. The Conference of Independent Data Protection Authorities sees P20 as an opportunity to create an open-source, controllable alternative to private operating systems; its chairman, Baden-Württemberg Commissioner for Data Protection Tobias Keber, said that if implemented correctly, the project could protect fundamental rights and strengthen digital sovereignty. On January 29, 2026, the conference unanimously called for a statutory basis for operating the data house, whose lines of responsibility remain unresolved. Beginning at the end of 2026, the Federal Criminal Police Office will migrate its data there, and the government emphasizes that P20 relies heavily on open source and European solutions, according to specialist reporting on the migration.

Yet the same government stated in its responses to the parliamentary inquiries that the analytical layer above the data house will follow a “modular and hybrid approach” that can expressly include commercially available products alongside in-house developments. This leaves open what P20 will ultimately become: the foundation of a sovereign, open-source solution – or a state-owned shell into which a private operating system continues to be inserted, whether American or European. No declaration of principle will decide that question. A procurement procedure will.

Three Paths, an Open Outcome

Three paths can be imagined from the current situation. The first is American lock-in: Through Bavaria’s framework agreement, Palantir remains the quasi-standard that other states can join without conducting their own procurement procedures, while the federal government’s product- and technology-neutral infrastructure described in Part 1 remains compatible with Palantir despite all reservations. The second is European fragmentation: ChapsVision and rola, along with national champions in France and Germany, divide the market under the banner of sovereignty; the applicable legal order shifts from Washington to Paris and Berlin – but the model remains the same: a private operating system, an embedded integration partner, and years of dependency. The third is emancipation: The P20 data house becomes a state-controlled, open-source foundation on which analytical services can be built that are subject neither to a foreign legal order nor to a publicly traded owner. Which path will prevail remains undecided. The facts to date – the framework agreement as a door opener, contracts without exit clauses, and P20’s expressly hybrid design – tilt toward the first two.

The finding of this second part is therefore sober. The capability is documented: the linking of millions of datasets, the speed in an emergency, and the reach of a corporation whose government business grew by 90 percent in a single quarter and whose market valuation embodies the bet that this dependency will endure. The intention behind it remains an open question. What is remarkable is not that private companies sell their software; that is their business. What is remarkable is that a state that has elevated digital sovereignty into a guiding principle delegates control over its most sensitive data infrastructure to publicly traded companies whose value depends on nobody ever being able to leave their platforms. Whether that operating system is American or European changes the legal order governing it and the names of those who profit from it – but not the structure. Whoever owns the operating system owns a switch inside the state.

Beneath both operating systems – the one for the financial markets and the one for the security state – lies a third layer that is rarely discussed: the cloud infrastructure on which they run. Aladdin is moving into Microsoft’s and Amazon’s data centers, while Palantir’s platform manages its own deployment through a control layer that reaches every environment. Who provides the computing power on which the operating systems of power run is the question the next part of this series will examine.

Published at apolut media on September 11, 2026

No advertisers, no sponsors, no political patrons – this publication is funded by its readers. Support independent geopolitical analysis from $5 per month and unlock all member-only content.

You are also welcome to support my work with a one-time donation.

About the Author

Michael Hollister
is a geopolitical analyst and investigative journalist. He served six years in the German military, including peacekeeping deployments in the Balkans (SFOR, KFOR), followed by 14 years in IT security management. His analysis draws on primary sources to examine European militarization, Western intervention policy, and shifting power dynamics across Asia. A particular focus of his work lies in Southeast Asia, where he investigates strategic dependencies, spheres of influence, and security architectures. Hollister combines operational insider perspective with uncompromising systemic critique – beyond opinion journalism. His work appears on his bilingual website (German/English) www.michael-hollister.com and in investigative outlets across the German-speaking world and the Anglosphere.

Understanding international politics takes more than following the news: independent analysis and background delivered directly by email

Sources

Court Rulings, Legislation, and Parliamentary Documents

1. German Bundestag, Printed Paper 21/7079 – Response by the German government to the minor interpellation submitted by The Left parliamentary group (Printed Paper 21/6752), purchase of software by the Federal Office for the Protection of the Constitution, July 08, 2026: PDF.
2. German Bundestag, Printed Paper 21/4923 – Response by the German government to the minor interpellation submitted by the AfD parliamentary group (Printed Paper 21/4599), possible use of software produced by Palantir by the Federal Police and the Federal Criminal Police Office, March 23, 2026: PDF.
3. German Bundestag, Printed Paper 21/5625 – Response by the German government to the minor interpellation submitted by the AfD parliamentary group (Printed Paper 21/5119), possible use of Palantir Gotham data-analysis software by German intelligence services, April 23, 2026: PDF.
4. German Federal Constitutional Court, judgment of the First Senate of February 16, 2023, 1 BvR 1547/19 and 1 BvR 2634/20, English version (use of the term “operating system”): Judgment.

Corporate and Financial Data

5. Palantir Technologies, second-quarter 2026 earnings release (Form 8-K, Exhibit 99.1, U.S. Securities and Exchange Commission): Earnings Release.
6. Palantir Technologies, Investor Relations: Investor Relations.
7. Macrotrends, “Palantir Technologies – Market Capitalization” (as of September 2026): Market-Capitalization Data.
8. TechTimes, “Palantir Q2 2026 Earnings and the Valuation Debate” (valuation, price-to-sales ratio, short position): Article.
9. Michael Burry, renewed short position against Palantir, September 2026; institutional reference: tipranks.com (search term: Michael Burry Palantir AI FOMO bubble September 2026).

Business Model and Lock-In – Forward Deployed Engineers

10. PitchGrade Research, “Palantir’s Business Model: Government Anchor, Commercial Optionality, and the AI Pivot” (investment-and-harvest cycle, security clearances as a switching barrier): Analysis.
11. Progressive Robot, “Forward-Deployed Engineering: How Enterprise AI Learns” (operational dependency, contractual safeguards, Amazon’s billion-dollar investment), September 02, 2026: Article.
12. Steve Banker, “Palantir and Forward Deployed Engineering: What Should We Believe?” Forbes (Charlie Gottdiener of Anaplan as a dissenting voice), July 10, 2026: Article.

ChapsVision, eFront, and rola Security Solutions

13. ChapsVision, “About Us” (Olivier Dellenbach, sale of eFront in 2019, CyberGov division, ArgonOS as a heterogeneous data operating system): Company Profile.
14. ChapsVision, press release, “ChapsVision and rola Security Solutions Enter into Partnership,” October 15, 2025: Press Release.
15. rola Security Solutions, press release on its strategic partnership with ChapsVision (division of responsibilities, Matthias Vianden): Press Release.
16. rola Security Solutions, company profile (wholly owned subsidiary of T-Systems International since 2014): Company Profile.
17. SVA System Vertrieb Alexander, press release, “SVA, rola and ChapsVision Launch Strategic Partnership for Security Agencies in Germany”: Press Release.
18. Police-IT, “T-Systems Acquires PIAV Preferred Candidate Rola Security Solutions” (2014 acquisition, technological sovereignty): Article.
19. Police-IT, “Rola Security Solutions: A State-Created Monopolist”: Article.
20. Intelligent Gov.tech, “ChapsVision Selected by French State to Deploy Sovereign Argonos Data Platform,” June 19, 2026: Article.

The P20 Program and Data Protection Oversight

21. German Federal Ministry of the Interior, “P20 Program (Police 20/20)”: Program Page.
22. German Federal Criminal Police Office, “The Police 20/20 Program”: Program Page.
23. North Rhine-Westphalia Commissioner for Data Protection and Freedom of Information, Data Protection Conference on the P20 police data house (resolution of January 29, 2026): Resolution.
24. Conference of Independent Data Protection Authorities, “Clear Statutory Rules for the P20 Police Data House” (Tobias Keber, unanimous resolution): Statement.
25. ChannelPartner, “Police to Rely on European IT in the Future” (P20, open source, Federal Criminal Police Office data migration beginning at the end of 2026): Article.

Earlier Analyses in This Series

26. Michael Hollister, “The Second Operating System of Power – How Palantir Connects and Divides Germany’s Security Agencies” (Private Operating Systems of Power, Part 1).
27. Michael Hollister, “The Operating System of Global Financial Markets – How BlackRock’s Aladdin Became Private Infrastructure” (BlackRock/Aladdin, Part 1).
28. Michael Hollister, “How BlackRock Profits from the Chancellor’s Policies” (BlackRock/Aladdin, Part 2).

© Michael Hollister – All rights reserved. Redistribution, publication or reuse of this text requires express written permission from the author. For licensing inquiries, please contact the author via www.michael-hollister.com.