NEW YORK, October 16, 2026 — Blackstone is positioned to commit approximately $9 billion to a $60 billion artificial intelligence semiconductor financing package benefiting Anthropic, marking one of the largest proposed private credit investments associated with the rapid expansion of AI computing infrastructure. The transaction, known as Project North Star, would finance specialized processors required to support Anthropic’s increasingly demanding AI workloads. Its structure brings together alternative asset managers, major investment banks, semiconductor suppliers and institutional lenders in an arrangement that could influence how future technology infrastructure projects are financed.

According to the Financial Times on October 16, Blackstone has been offered a double-digit yield to participate in the transaction, reflecting the substantial financing requirements and credit risks associated with large-scale AI infrastructure deployment. The proposed investment represents half of an $18 billion junior debt tranche within the overall package. The remaining $42 billion would consist of senior secured financing supported by Broadcom, which is involved in supplying custom processors developed for Google’s artificial intelligence computing ecosystem.

The transaction has attracted considerable attention because it combines the scale of investment typically associated with major infrastructure projects with the faster commercial development cycles of the technology industry. Unlike conventional financing for long-lived assets such as airports, utilities or telecommunications networks, lending against AI semiconductors introduces additional uncertainty surrounding equipment obsolescence, future utilization, customer demand and the recoverable value of specialized hardware.

Blackstone’s anticipated participation also demonstrates how private credit managers are becoming increasingly important intermediaries in financing the physical infrastructure supporting AI development. Rather than relying exclusively on corporate balance sheets, equity issuance or conventional bank loans, technology companies and their suppliers are using structured financing arrangements to fund computing capacity without necessarily paying the entire equipment cost upfront.

Project North Star is structured around two principal categories of debt. The $42 billion senior secured tranche is expected to receive credit support from Broadcom, providing an additional source of protection for lenders. The $18 billion junior tranche carries greater exposure to Anthropic’s underlying financial performance and the economics of its computing arrangements. Blackstone is expected to anchor the junior financing with its approximately $9 billion participation, while the balance would be distributed to additional investors.

The distinction between senior and junior financing is central to the transaction’s economics. Senior secured lenders generally receive priority in the repayment hierarchy and may benefit from contractual protections, collateral rights and guarantees or other credit support. Junior investors accept a lower position in the capital structure, potentially exposing them to larger losses if the transaction encounters financial difficulties. In exchange, they typically demand higher interest rates or additional economic compensation.

Blackstone’s reported yield exceeding 10% illustrates this relationship between expected returns and underlying risk. For institutional investors, a double-digit return may provide attractive income relative to many conventional investment-grade corporate bonds. However, that additional return compensates lenders for uncertainties that extend beyond ordinary corporate credit exposure, including the commercial performance of Anthropic, the financial structure of the equipment purchases and the evolving market for AI computing capacity.

The proposed financing follows an earlier transaction known as Project Big Sky, which helped establish a model for funding expensive AI processors through specialized lending structures. That arrangement included approximately $4.4 billion in junior financing carrying an interest rate of about 8.5%, according to the Financial Times. The higher reported return associated with Project North Star suggests that investors are seeking additional compensation as financing volumes increase and concerns about the sustainability of AI infrastructure investment intensify.

The difference between the two reported borrowing costs is notable. A rise from approximately 8.5% to a yield exceeding 10% represents an increase of more than 1.5 percentage points. Applied to billions of dollars in financing, relatively modest changes in required returns can materially affect the economics of equipment leasing and infrastructure development. The comparison does not establish that the two transactions carry identical credit risks, maturities or contractual protections, but it illustrates the increasing importance of financing costs in AI infrastructure decisions.

The financing would support the acquisition of computing equipment based on Google’s tensor processing units, specialized semiconductors designed to accelerate machine-learning workloads. The Financial Times reported that Project North Star is associated with approximately 1.5 gigawatts of computing capacity. Such a deployment represents a substantial industrial investment, requiring not only processors but also supporting networking equipment, power delivery, cooling infrastructure and suitable data-center facilities.

Broadcom occupies a strategically important position in the proposed arrangement. The semiconductor company works with Google on customized AI processors and associated networking technology. In an April 6 regulatory filing, Broadcom disclosed an expanded collaboration with Google and Anthropic under which Anthropic would begin accessing approximately 3.5 gigawatts of next-generation TPU-based AI computing capacity through Broadcom from 2027. That broader commitment provides commercial context for the current financing initiative, although it should not be treated as identical to the 1.5-gigawatt capacity associated with Project North Star.

The same filing stated that consumption of the expanded computing capacity would depend on Anthropic’s continued commercial success. This qualification is important for institutional lenders assessing the transaction. Semiconductor financing relies not only on the availability and technical performance of the equipment but also on the borrower’s ability to generate sufficient revenue from the computing resources financed by the transaction.

For Anthropic, access to additional processors could strengthen its ability to train and operate advanced artificial intelligence models while managing the timing of infrastructure expenditures. AI developers face substantial upfront costs because sophisticated models require large computing clusters, extensive networking infrastructure and continuing operational investment. Leasing arrangements and structured equipment financing can distribute these costs over time, potentially aligning payments more closely with commercial use of the infrastructure.

However, financing equipment over several years introduces a financial obligation that persists even if business conditions change. Anthropic’s ability to support the relevant payments will depend on demand for its AI services, customer retention, enterprise adoption, pricing power and the efficiency with which its computing resources generate revenue. A slower expansion in AI usage or intensified competition could reduce the financial flexibility available to meet those obligations.

The presence of Broadcom’s credit support in the senior tranche is designed to address part of this risk. Bloomberg previously reported that Broadcom would provide residual-value support for the $42 billion senior secured financing. Such protection can improve the risk profile of debt investors by providing contractual backing associated with the equipment’s remaining value. The precise effectiveness of that protection depends on the finalized documents, the scope of Broadcom’s commitments and the circumstances under which the support becomes available.

Residual-value considerations are particularly important for AI processors because technological developments can rapidly change the economics of existing hardware. New generations of chips may deliver more computing power per unit of electricity or offer improved performance for emerging AI applications. If newer equipment becomes substantially more efficient, older processors could experience declining market values even while remaining technically operational.

That challenge distinguishes semiconductor-backed lending from many traditional forms of equipment finance. Aircraft, commercial vehicles and industrial machinery often have established secondary markets that help lenders estimate recovery values. Highly specialized AI processors may have narrower resale opportunities, potentially greater dependence on compatible software and networking systems, and uncertain values once a new hardware generation becomes commercially available. These characteristics make collateral valuation and contractual protection important components of the transaction.

Project North Star also illustrates the expanding role of Wall Street banks in distributing large technology financing packages. Bloomberg reported earlier in October that Bank of America, Citigroup and Morgan Stanley were among the institutions involved in arranging the senior secured financing. Their participation highlights the importance of bank syndication, through which large credit exposures can be distributed among multiple lenders rather than concentrated entirely on the balance sheets of the original arrangers.

For the participating banks, the transaction presents opportunities to generate underwriting and arrangement fees while strengthening relationships with major technology companies and institutional asset managers. Nevertheless, the banks must assess investor appetite, contractual terms, credit support and the amount of financing that can be successfully distributed. Their exposure ultimately depends on the amounts retained and the underwriting commitments established in the final financing documents.

The junior tranche presents a different distribution challenge. Investors considering subordinated exposure must evaluate Anthropic’s financial position more directly because the financing does not benefit from the same level of Broadcom support as the senior portion. According to Financial Times reporting, Anthropic received a BB-plus credit assessment from Kroll Bond Rating Agency in June, placing it one notch below investment grade. Although this represents the highest category of speculative-grade credit, it still indicates meaningful exposure to adverse business conditions.

The relationship between Anthropic’s credit profile and the cost of financing could become increasingly important as the company considers future capital-market activity. Greater financial disclosure, a longer operating record and improved credit quality could broaden the range of potential lenders. Conversely, deteriorating financial performance or weaker demand for AI infrastructure could raise borrowing costs, reduce investor participation or require additional financial protections.

Reports have also linked the proposed financing structure to Anthropic’s potential public-market ambitions. A future stock-market listing could expand financial transparency and influence the company’s access to institutional credit. However, a public offering would not automatically guarantee a higher credit rating or lower financing costs. Rating agencies and investors would continue to evaluate liquidity, profitability, debt obligations, customer concentration and the sustainability of cash generation.

Blackstone’s involvement is consistent with its broader investment strategy across AI-related assets. Bloomberg reported on October 15 that the alternative asset manager had assembled a substantial data-center investment portfolio and was expanding its activities into AI computing infrastructure and related businesses. Its participation in Project North Star would add another major credit commitment to an investment strategy spanning physical infrastructure, technology services and financing arrangements.

The firm has also developed commercial relationships with Anthropic beyond semiconductor lending. In May, Anthropic, Blackstone, Hellman & Friedman and Goldman Sachs announced the formation of an enterprise AI services business intended to help companies integrate Claude into their operations. That initiative illustrates the multiple ways large alternative asset managers are participating in AI adoption, from providing financial capital and owning infrastructure to supporting enterprise deployment.

For Blackstone’s investors, the proposed chip financing could provide additional income exposure to AI infrastructure without relying exclusively on equity appreciation. Debt investments generate contractual payment obligations and can offer different risk-return characteristics from ownership interests. Nevertheless, the underlying economic exposures remain connected to the same technological and commercial developments affecting AI infrastructure valuations.

The scale of Blackstone’s potential commitment also raises questions about concentration risk. A $9 billion position would represent approximately 15% of the overall Project North Star financing and 50% of the junior tranche. Bloomberg has reported that the proposed commitment would come from various Blackstone-managed funds, meaning the economic exposure would not necessarily reside within a single investment vehicle. The ultimate distribution across funds and clients will depend on transaction documentation and investment allocations.

For institutional investors participating through private credit funds, insurance portfolios or other managed vehicles, exposure to AI infrastructure may be less visible than direct ownership of publicly traded technology shares. The underlying investments can nevertheless be sensitive to comparable developments, including changes in AI demand, semiconductor economics, energy availability, interest rates and corporate spending priorities.

Market conditions have become an increasingly relevant consideration. Financial Times reporting on October 15 described weakening investor sentiment toward certain data-center-related bonds as concerns intensified about leverage and the profitability of AI infrastructure projects. The broader market environment suggests that institutional lenders are distinguishing more carefully between transactions supported by strong contractual cash flows and those requiring optimistic assumptions about future utilization or technology demand.

Project North Star’s successful execution would therefore carry significance beyond Anthropic and Blackstone. A completed financing at the proposed scale could demonstrate that institutional markets remain capable of absorbing exceptionally large AI-related credit transactions, even as investors become more selective about pricing, collateral and counterparty exposure. Conversely, difficulties distributing the debt could indicate that the amount of capital available for certain AI infrastructure projects is becoming more constrained.

The transaction is also relevant to the competitive relationship between custom AI processors and the broader semiconductor market. Google’s TPUs, developed with Broadcom’s involvement, offer an alternative architecture for major AI workloads. Financing arrangements capable of supporting large custom-chip deployments could influence how AI developers make long-term infrastructure commitments. For lenders, however, the primary consideration remains whether the equipment and associated contracts can produce sufficient financial returns to support the capital invested.

Several unresolved questions will determine the final economics of Project North Star. These include the contractual maturity of each debt tranche, the amortization schedule, the interest-rate structure, the extent of collateral coverage, the precise terms of Broadcom’s support and the protections available to junior creditors. Investors will also examine any restrictions on additional borrowing, provisions governing equipment performance and circumstances that could trigger early repayment or restructuring.

Another consideration is the timing of equipment deployment relative to the commencement of financing obligations. Construction schedules, utility connections, semiconductor deliveries and data-center commissioning can affect when computing assets become commercially productive. If financing costs accumulate before the equipment generates adequate revenue, the transaction may require additional liquidity support or financial reserves to manage the interval.

The availability of electricity and supporting infrastructure remains a practical constraint on deploying computing capacity measured in gigawatts. A large processor installation requires substantial power supply, cooling and network connections. Although Project North Star focuses on semiconductor financing rather than the entire construction program, delays in related infrastructure could influence equipment utilization and the economics of the associated leasing contracts.

For lenders evaluating the proposal, a central issue will be whether projected cash flows remain sufficient under less favorable commercial assumptions. Stress scenarios may involve slower customer growth, lower computing prices, unexpected operating costs, delayed hardware deployment or accelerated technological replacement. The relative resilience of the senior and junior tranches under those scenarios will influence institutional demand and the returns investors require.

Blackstone’s willingness to participate at the reported yield indicates continued interest in supplying capital to AI-related infrastructure projects where the expected compensation is considered sufficient. It does not eliminate the possibility that financial terms could change before completion. Large syndicated transactions frequently require adjustments to pricing, allocation and investor protections as market conditions and lender commitments develop.

As of October 16, the reported $9 billion Blackstone commitment and the broader $60 billion financing structure represent an advanced financing initiative rather than evidence that the entire transaction has been funded and completed. The final allocation of debt, pricing and completion timetable remain important matters for market participants to monitor. Investors will also be watching how the proposed junior financing is distributed beyond Blackstone and whether the senior secured tranche attracts sufficient demand.

The broader implication is that the AI infrastructure expansion is increasingly becoming a major credit-market financing challenge rather than simply an equity-market investment theme. Developing advanced AI systems requires access to specialized hardware, substantial electricity supplies and long-term capital. Private credit firms and investment banks are emerging as important providers of that capital, while semiconductor suppliers are taking on increasingly complex roles in supporting equipment financing.

Project North Star brings those developments together in a single proposed transaction. For Anthropic, it could help secure additional computing resources essential to expanding its AI business. For Broadcom, it could support the commercial adoption of custom semiconductor technology. For Blackstone and other institutional lenders, it represents an opportunity to earn substantial financing returns while assuming exposure to one of the most capital-intensive areas of the technology economy.

Ultimately, the significance of the transaction will depend on more than its headline size. The decisive measures will be the financing terms accepted by investors, the durability of the underlying cash flows, the effectiveness of contractual credit protections and the ability of the financed computing equipment to remain economically productive. If successfully completed, Project North Star could provide a prominent reference point for future semiconductor-backed lending and the continued expansion of private capital into artificial intelligence infrastructure.