The Machine That Prices the World

BlackRock’s Aladdin platform prices more than $25 trillion in financial assets, advises the central banks that are supposed to regulate it, and now sits inside the consortiums buying Panama’s ports, building the data centers that will run artificial intelligence, and structuring Ukraine’s reconstruction.

A Workstation Between a Coffee Maker and a Refrigerator

In 1988, in a single room with eight people, a bond trader named Larry Fink and seven partners built a risk-analysis system on a Sun Microsystems workstation that sat, by every available account of the firm’s own early history, wedged between an office coffee maker and a refrigerator. The system existed to do one narrow thing: price mortgage-backed securities accurately enough that the firm trading them would not be blindsided by the kind of collateral collapse that had just ended Fink’s career at First Boston, where a bad bet on interest rates had cost the bank $100 million and cost him his job. The machine they built to keep that from happening again was eventually given a name, Aladdin, an acronym for asset, liability, debt and derivative investment network. By 1999 BlackRock had begun licensing it to outside clients.

Aladdin no longer prices mortgage bonds for one trading desk. As of the most recent disclosures available to the firm’s clients and counted across more than a thousand institutions worldwide, the platform now sits underneath approximately $25 trillion in assets, a figure equivalent, depending on how the global total is counted, to somewhere between a twentieth and a fifteenth of all financial assets on earth. It does this with a client list that includes BlackRock’s own $14 trillion under direct management, more than half of the world’s ten largest insurers, sovereign and public pension funds across multiple continents, and corporate treasuries at Apple, Microsoft and Alphabet, three companies that compete with each other and with BlackRock’s own holdings but that nonetheless run their cash through the same risk engine. The platform has expanded onto both Microsoft Azure and, as of a partnership announced in December 2025, Amazon Web Services, with full availability across both clouds expected in the second half of this year. A generative AI layer called Aladdin Copilot, built on Azure’s OpenAI infrastructure, now sits on top of that, answering natural-language questions about portfolios that collectively dwarf the GDP of every country except the United States and China combined.

None of this happened by conquest. It happened because every institution that adopted Aladdin made, in isolation, a rational decision: better risk modeling, lower operating cost, a single data language across an entire portfolio. The aggregate of a thousand rational decisions is a financial nervous system with one company’s hands on the wiring.

A Twentieth of Everything, One Company’s Eyes on All of It

The conflict embedded in that arrangement is not hidden. BlackRock discloses it in its own fund prospectuses, in language that has appeared in SEC filings for years: BlackRock has a financial incentive to recommend BlackRock products, and BlackRock’s stewardship and analytics functions operate alongside its asset-management functions inside the same corporate structure. The firm’s public position is that internal walls, conflict committees and disclosed methodology manage this. The structural fact underneath the disclosure does not change. The same company that builds the risk model competitors are required to trust also manages money that competes against them, votes shares in nearly every public company of consequence on the planet, and increasingly decides which data points, from climate exposure to reputational risk, get fed into the model in the first place. In July 2025, Aladdin absorbed a corporate reputational-risk dataset from the Swiss firm RepRisk covering more than 400,000 companies, adding yet another layer of judgment, made by BlackRock, about what counts as risk, to a system that more than a thousand other institutions now treat as a neutral utility.

A risk model is never neutral. It is a set of decisions about what to measure, expressed in code instead of in a memo, and the company writing that code is also a market participant with a balance sheet at stake in the outcome.

The Regulator Becomes the Regulated

The clearest demonstration of what that means came in 2008, when the United States government, lacking the legal authority to buy toxic mortgage assets directly, turned to BlackRock to manage three Federal Reserve special-purpose vehicles known as Maiden Lane, Maiden Lane II and Maiden Lane III. Maiden Lane absorbed roughly $30 billion in distressed assets to induce JPMorgan into buying the surviving pieces of Bear Stearns. Maiden Lane II and III absorbed the mortgage-backed securities and collateralized debt obligations that American International Group could not make good on to the banks it had insured. The same year, the federal government brought BlackRock in to evaluate the toxic asset books of Fannie Mae and Freddie Mac after seizing both. The contracts were awarded without competitive bidding.

Twelve years later the arrangement repeated at greater scale and in plain daylight. In March 2020, the Federal Reserve hired BlackRock’s Financial Markets Advisory unit to manage a corporate-bond purchase program launched in response to the pandemic shock, and BlackRock began buying Treasury and agency mortgage-backed securities on the Fed’s behalf days later. Documents the New York Fed eventually published showed that BlackRock employees staffing the central bank’s account were permitted to resume trading on what they had learned while advising it after a cooling-off period of two weeks, during which they were still allowed to share what the financial press at the time described as general market views with BlackRock’s own clients. Bloomberg reported plainly that the arrangement put the firm on both sides of the trade: managing the central bank’s purchases while remaining free, after a fortnight, to buy some of its own exchange-traded funds on the central bank’s behalf and trade around the very market the central bank was moving with its money.

The official framing was that no other institution had the operational capacity to execute a rescue program of that size on short notice, and that framing is not false. It is also not the whole mechanism. The whole mechanism is a private company that prices and trades the assets a crisis devalues being handed, twice in twelve years, sole control of the public response to that devaluation, with no public vote, no competitive process, and a cooling-off period measured in days rather than years.

Advisor, Structurer, Manager: One Client, Three Jobs

The same triple role appears, with the conflict simply relocated from Washington to Kyiv, in BlackRock’s relationship with Ukraine. In November 2022, nine months into the full-scale Russian invasion, Ukraine’s Ministry of Economy appointed BlackRock’s Financial Markets Advisory group, on a pro bono basis, to design the structure of what became the Ukraine Development Fund, intended to blend public guarantees from Western governments with private institutional capital to finance reconstruction. By January 2026, following a Coalition of the Willing meeting in Paris, US Special Envoy Steve Witkoff was describing the arrangement, now branded the Ukraine Prosperity Plan and targeting up to $800 billion in mobilized investment over a decade, as unprecedented cooperation between Washington and a single asset manager on the rebuilding of a country still at war.

The advisory work was pro bono. The structuring work was not incidental. A firm that designs the rules by which a war-shattered economy’s infrastructure, energy and mineral assets get offered to outside capital is, by the nature of that design work, positioned to identify which of those assets it or its institutional clients should then buy, on terms it helped write, in a negotiating environment where the seller’s bargaining power has been reduced by the war itself to something close to zero.

That reduction has a denomination the prospectus does not carry. A country’s leverage collapses because its power stations are rubble, its grid is dark, and its working-age men are at the front or in the ground, and the further the seller’s position falls, the better the terms the buyer is able to write. The hundreds of thousands of Ukrainian dead are not a line in the Ukraine Development Fund’s documentation. They are the reason the entry is priced the way it is. The discount is the war, and the war is the dead, and neither appears in the column where the returns are totaled.

Western governments provide the guarantees that de-risk the entry. Ukraine provides the assets and the long-term revenue commitments. The capital that BlackRock helps mobilize collects the inflation-protected returns on both. The shape of that arrangement, public guarantee absorbing the downside while private capital captures the upside on assets priced under duress, is not new in the history of capital meeting a state with no leverage left to negotiate with. What is new is having the same firm sit on every side of the table at once: advisor to the government doing the selling, structurer of the fund doing the buying, and prospective manager of the capital that flows through it.

The Canal

Geography makes the same structure visible without requiring a war to explain it. In March 2025, a consortium led by BlackRock, alongside its infrastructure arm Global Infrastructure Partners and the Mediterranean Shipping Company’s terminal subsidiary, agreed to buy CK Hutchison’s 90 percent stake in the Panama Ports Company, the concession holder running the Balboa and Cristobal terminals at either end of the Panama Canal, in a deal covering 43 ports across 23 countries and valued near $23 billion. The Trump administration, which had spent months describing Chinese influence over the canal as a national security threat and had openly discussed reclaiming it, treated the sale as a geopolitical win before the ink was dry.

Beijing did not. China’s State Administration for Market Regulation opened an antitrust review of a transaction involving a Hong Kong company selling assets that did not include a single Chinese port, a move widely read in Hong Kong and in Washington alike as political leverage rather than competition policy. Through the back half of 2025, Chinese state media accused CK Hutchison of capitulating to American pressure, and Beijing pushed for the state-owned shipping line Cosco to be added to the buying consortium, first at a minority stake and then, by December, demanding outright majority control as a condition of letting the deal proceed at all. BlackRock and MSC reportedly considered walking away rather than concede a Chinese state enterprise control of the position. The standoff ended not in a boardroom but in a Panama courtroom: in February 2026, Panama’s government annulled CK Hutchison’s port concessions outright following a Supreme Court ruling, and handed interim operational control of Balboa and Cristobal to Maersk and MSC while CK Hutchison opened arbitration proceedings against the Panamanian state.

Whatever the final ownership structure that emerges from that arbitration, the episode already states the argument plainly. A chokepoint carrying roughly three percent of global seaborne trade became, over eleven months, the terrain on which Washington and Beijing fought a proxy contest for control, and the entity initially positioned to hold the winning hand was not a navy or a foreign ministry. It was an asset manager’s infrastructure fund.

Building the Machine That Will Run the Machine

If Aladdin is the operating system for the world’s existing capital, BlackRock’s most consequential recent move has been securing a position inside the construction of the operating system for the next one. In September 2024, BlackRock, its Global Infrastructure Partners unit, Microsoft and MGX, a technology investment vehicle created in Abu Dhabi that same year, launched what was first called the Global AI Infrastructure Investment Partnership and is now known simply as the AI Infrastructure Partnership. Nvidia joined as a technical advisor, and by March 2025 Elon Musk’s xAI had joined as a partner alongside it. The Kuwait Investment Authority and Singapore’s Temasek followed. The partnership’s stated goal is to mobilize $30 billion in private equity capital from investors, asset owners and corporations, leveraging that base toward as much as $100 billion in total investment once debt financing is layered on top, all of it directed at the data centers and energy infrastructure that generative AI requires to function.

In October 2025 the partnership agreed the largest digital-infrastructure transaction in history, a $40 billion acquisition of Aligned Data Centers, a company operating fifty campuses and more than five gigawatts of capacity across the Americas, bought from Macquarie Asset Management by a group that included BlackRock’s infrastructure arm, MGX, Microsoft, Nvidia and xAI together, a deal expected to close in 2026. On BlackRock’s fourth-quarter 2025 earnings call, Fink told investors the partnership had already raised more than $12.5 billion from its founders and clients and called it a multi-trillion-dollar long-term opportunity.

Lay the two facts beside each other without commentary. The same firm whose risk software already prices more than $25 trillion in financial assets is now an equity owner, alongside the chip company that supplies the processors and a Gulf sovereign vehicle that supplies a portion of the capital, of the physical plant on which the artificial intelligence systems of the next decade will run, including, eventually, the generative AI layer sitting on top of Aladdin itself. The financial complex and the technological complex are not converging by accident. They are converging because the same handful of firms are building both sides of the convergence, and the public utilities regulators who might once have asked whether that concentration serves anyone but the firms involved have not yet finished deciding what question to ask.

The States That Said No

The clearest organized resistance to BlackRock’s reach has come not from the political left, where most of the firm’s critics over its asset-stewardship role have historically sat, but from a coalition of Republican-led states objecting to the opposite complaint: that BlackRock’s environmental and governance commitments amounted to using other people’s retirement savings to enforce a climate policy no legislature had voted for. Louisiana’s state treasurer pulled $794 million in 2022, telling the firm directly that its position “would destroy Louisiana’s economy.” Missouri pulled $500 million. Mississippi issued a cease-and-desist order. Texas, in 2024, moved to divest $8.5 billion, a sum independent actuarial analysis suggested could cost the state’s own pension beneficiaries billions in lost returns over a decade, a cost the divesting states have generally been willing to absorb as the price of the principle.

In November 2024 that political objection became a federal antitrust case. Texas and eleven other states sued BlackRock, Vanguard and State Street under the Sherman and Clayton Acts, alleging the three firms used their combined ownership stakes across nine major American coal producers to pressure those companies into cutting output in service of climate commitments made through groups like the Net Zero Asset Managers Initiative, a campaign the states say coincided with thermal coal production falling sharply between 2019 and 2022 even as coal prices rose. In August 2025, US District Judge Jeremy Kernodle in the Eastern District of Texas declined to dismiss the bulk of the case, finding the states had assembled enough circumstantial evidence of parallel conduct, shared timing and shared motive to let the conspiracy claim proceed past the pleading stage, while noting the states would still need to prove it. The Department of Justice and Federal Trade Commission filed a joint statement that May supporting the theory that common ownership could, in principle, implicate antitrust law even without direct evidence of coordination, a position neither agency had been willing to commit to before. Vanguard settled in February 2026 for $29.5 million and an agreement to keep its American business out of future climate-stewardship coalitions, while State Street called the underlying allegations baseless and BlackRock has continued to fight the case in court.

Groupthink, by Design

The institutional concern that predates the political one is more mechanical and, in its way, more frightening: what happens when the financial system’s principal risk gauge is the same gauge for nearly everyone holding the asset being measured. In January 2021, the Los Angeles County Employees Retirement Association, a $58 billion pension fund, passed over Aladdin’s risk-analytics offering in favor of a competitor, MSCI, citing in its own board materials the danger of what it called groupthink, the prospect that hundreds of institutions reading identical risk signals would behave identically during a market shock, turning a single platform’s blind spot into everyone’s blind spot at once. The Financial Times, reporting on the same concern in 2021, quoted a former chief executive of a $476 billion fund manager warning that anything resembling an oligopoly in risk management would be especially dangerous if any weakness surfaced in the underlying system.

That warning has not been tested by a true stress event since Aladdin grew to its current scale, and the absence of a test is not evidence the risk is smaller than it looks. It is evidence only that the test has not yet come.

What Beijing Will Not Allow

China’s answer to all of this has been to wall its own financial data off rather than negotiate access to it. New measures from the People’s Bank of China took effect on June 30, 2025, governing how financial institutions under its oversight, including banks, insurers, asset managers and leasing companies, can process and store data generated inside the PBOC’s own regulatory domain, with state secrets carved out entirely from any foreign system’s reach. The rule does not name BlackRock or Aladdin. It does not need to. A regulator that has spent three years using merger review as leverage to force a state shipping company into a foreign-led ports consortium has already shown what it thinks of letting an American firm’s risk software sit inside the institutions that manage the People’s Republic’s own financial exposure. Sovereignty over data and sovereignty over chokepoints are, to Beijing, the same fight conducted on two different fronts.

Tokenizing What Is Left

The newest phase of BlackRock’s expansion runs through the part of the financial system that has so far stayed outside Aladdin’s reach almost entirely: the retirement savings of ordinary households. In his 2026 annual letter to shareholders, Fink argued that tokenization, converting equities, bonds and funds into blockchain-based digital tokens, will do for investing what the internet did for information, framing BlackRock’s own tokenized money market fund, BUIDL, and a reported $65 billion in stablecoin reserves linked to the firm as the early proof of the model. The letter arrived as 2026 became BlackRock’s first full year operating as what Fink called a unified platform following its absorption of Global Infrastructure Partners, the private-credit firm HPS, and the private-markets data provider Preqin, three acquisitions that together give one company the infrastructure money, the private lending relationships and the data standard for an entire asset class that public markets cannot independently see into.

That consolidation lands at the same moment the US Department of Labor has been reviewing rules that would open employer-sponsored retirement plans to private-market and alternative investments for the first time at scale, a market Fink has openly framed as BlackRock’s next growth opportunity and one where Aladdin and the eFront platform, not an independent public exchange, would supply the benchmarks and analytics regulators and plan administrators rely on to judge whether the investments are fairly priced at all. BlackRock entered 2026 managing a record $14 trillion after taking in $698 billion of net new client money in 2025 alone, the strongest year of inflows in the firm’s history.

A firm that began as eight people pricing mortgage bonds on a borrowed workstation now sits inside the Federal Reserve’s emergency response, a war-torn nation’s reconstruction fund, a contested shipping chokepoint between two nuclear powers, the equity stack of the data centers that will train the next generation of artificial intelligence, and the benchmark layer that may soon decide how tens of millions of retirement accounts get priced into markets those savers cannot see. Each of those positions was acquired through a contract, a court filing, a regulatory comment period or a shareholder letter, all of them public, all of them legal, none of them requiring a conspiracy to explain. That is the part that should hold the attention: not that a rule was broken, but that none had to be.

The question the record raises and does not answer is simpler than the word globalism makes it sound. When the same machine prices the risk, advises the buyer, structures the deal, and now builds the hardware the next version of itself will run on, there is no longer an outside to the loop. There is no independent party left to check whether the numbers still mean what they say, because the party that would check is the party that wrote them.