Broadcom is negotiating with lenders to raise more than $60 billion in debt as part of a massive financing arrangement aimed at supporting artificial-intelligence chip deployments for companies including Anthropic and potentially other AI firms. The proposed package could ultimately reach as much as $100 billion, making it one of the largest debt-financing structures tied to the rapidly expanding AI infrastructure market.
The financing discussions highlight how the AI boom is increasingly being funded through debt and private-credit markets rather than solely through corporate cash flows and equity investment. Broadcom is considering a senior-secured tranche of roughly $60 billion to $70 billion alongside about $30 billion of junior debt. The semiconductor company could guarantee part of the senior debt, while investment firms Blackstone and Apollo Global Management are among the potential financing participants.
Broadcom Seeks More Than $60 Billion For AI Chips
Broadcom is in talks with lenders to raise more than $60 billion for an AI chip financing arrangement, according to Bloomberg reporting cited by Reuters. The financing could be structured through a special-purpose vehicle, similar to an earlier Broadcom-backed AI infrastructure transaction.
The structure under discussion could be substantially larger than the headline $60 billion figure.
| Proposed Broadcom AI Financing | Amount |
|---|---|
| Senior-secured debt | $60 billion–$70 billion |
| Junior debt | About $30 billion |
| Potential total financing | Up to $100 billion |
| Minimum reported financing | More than $60 billion |
| Potential senior-debt guarantee | Broadcom |
| Potential financial participants | Blackstone, Apollo |
| Primary focus | AI chips and infrastructure |
The exact size and terms of the transaction have not been finalized. The financing is still under negotiation, meaning the eventual amount, structure and participating investors could change.
Why The Financing Is So Large
The scale reflects the enormous capital requirements of modern AI infrastructure.
Training and operating advanced AI models requires large clusters of accelerators, networking equipment, storage systems and data centers. AI companies are therefore committing billions of dollars to computing capacity before they can fully monetize the resulting services.
Broadcom sits in an important position within this ecosystem because it designs custom AI accelerators and supplies networking and connectivity technologies used in large data centers.
The company is therefore increasingly involved not only in selling chips but also in financing the infrastructure needed to deploy them.
The Deal Could Reach $100 Billion
The proposed structure contains two major layers of debt.
The senior-secured component could range between $60 billion and $70 billion. Senior debt receives priority over junior obligations in the event of a default or bankruptcy, making it generally less risky for lenders.
A separate junior tranche of approximately $30 billion could bring the overall financing package toward $100 billion.
| Debt Type | Potential Amount | Position |
|---|---|---|
| Senior-secured | $60B–$70B | Higher repayment priority |
| Junior | ~$30B | Lower repayment priority |
| Total | Up to $100B | Combined financing |
The structure allows investors with different risk appetites to participate in the same AI infrastructure opportunity.
For Broadcom, guaranteeing part of the senior debt could help lower financing costs and make the transaction more attractive to lenders, although it would also expose the company to additional financial obligations if the underlying projects failed to generate expected cash flows.
Anthropic Is A Major Beneficiary
Anthropic is expected to be one of the major beneficiaries of Broadcom’s AI infrastructure financing efforts.
In June, Broadcom partnered with Apollo and Blackstone to establish the AI XPV Platform, a financing vehicle designed to enable more than 20 gigawatts of computing capacity using Broadcom’s custom AI chips and networking technology by 2028.
The platform launched with a $35 billion transaction supporting more than 1 gigawatt of computing capacity for Anthropic.
| Broadcom AI XPV Platform | Figure |
|---|---|
| Initial financing | $35 billion |
| Initial compute capacity | More than 1 GW |
| Longer-term target | More than 20 GW |
| Target year | 2028 |
| Initial major beneficiary | Anthropic |
| Key investors | Apollo, Blackstone |
| Chip/network supplier | Broadcom |
The latest financing discussions appear to build on this broader model of using private capital to fund AI computing infrastructure.
From $35 Billion To Potentially $100 Billion
The difference between the earlier $35 billion platform and the latest potential $100 billion financing illustrates how quickly AI infrastructure financing is scaling.
The proposed new transaction could be nearly three times the size of the initial AI XPV transaction if it reaches its upper end.
| Financing | Size | Relative Scale |
|---|---|---|
| AI XPV initial transaction | $35B | 1x |
| Proposed minimum new financing | $60B+ | 1.7x+ |
| Proposed senior tranche | $60B–$70B | 1.7x–2x |
| Potential total package | $100B | 2.9x |
The comparison does not mean the transactions are identical. The $35 billion platform is an established financing arrangement, while the $100 billion figure remains a potential upper limit under discussion.
OpenAI Could Also Be Part Of The Broader Strategy
The latest reporting specifically identifies Anthropic and other companies as potential beneficiaries. OpenAI may also be connected to Broadcom’s broader custom-chip strategy.
OpenAI and Broadcom announced a partnership in October 2025 to develop and deploy 10 gigawatts of custom AI accelerators. OpenAI said the systems would combine its own accelerator designs with Broadcom’s semiconductor and networking technology, with deployments targeted to begin in the second half of 2026 and continue through 2029.
| OpenAI-Broadcom Partnership | Detail |
|---|---|
| Custom AI accelerator capacity | 10 GW |
| Partnership announced | October 2025 |
| Deployment start target | Second half of 2026 |
| Deployment completion target | End of 2029 |
| OpenAI role | Accelerator and system design |
| Broadcom role | Development and deployment |
However, the latest debt-financing reports do not establish that OpenAI is definitely a beneficiary of the proposed financing. Reports describe Anthropic and other AI companies as targets, while the exact list of customers has not been publicly disclosed.
Broadcom Expects AI Chip Revenue To Surge
Broadcom’s willingness to participate in very large infrastructure financing deals is connected to its growing AI-chip business.
The company expects AI-related chip revenue to exceed $100 billion in 2027, according to reporting on its business outlook. Anthropic is reportedly expected to account for more than 40% of that amount.
| Broadcom AI Business Outlook | Figure |
|---|---|
| Expected AI-chip revenue in 2027 | More than $100 billion |
| Reported potential Anthropic contribution | More than 40% |
| Potential Anthropic-linked revenue at 40% | More than $40 billion |
| AI revenue significance | Major growth driver |
A 40% share of $100 billion would imply more than $40 billion in revenue, although the actual contribution could be higher because the reported figure is described as “more than 40%.”
The numbers demonstrate why securing long-term AI infrastructure relationships is strategically important for Broadcom.
Broadcom Is Competing In Custom AI Chips
The AI-chip market is increasingly moving beyond Nvidia’s general-purpose GPUs.
Large technology companies are developing custom accelerators tailored to their own workloads. Google has its Tensor Processing Units, while OpenAI is developing its own accelerator architecture with Broadcom.
Broadcom plays a particularly important role because it can help hyperscalers and AI labs design custom silicon rather than relying exclusively on off-the-shelf GPUs.
| Company/AI Lab | Custom AI Hardware Strategy |
|---|---|
| TPU family | |
| OpenAI | Custom accelerator with Broadcom |
| Anthropic | Increasing use of custom/alternative compute |
| Meta | Custom AI silicon initiatives |
| Broadcom | Custom accelerators and networking |
| Nvidia | GPUs and broader AI accelerator platform |
This market shift could create a significant opportunity for Broadcom as AI companies seek to optimize performance, energy consumption and cost for specific workloads.
AI Infrastructure Is Becoming A Financing Business
The Broadcom transaction is part of a broader transformation in the AI economy.
AI infrastructure requires so much capital that traditional corporate financing is increasingly being supplemented by private credit, project finance, special-purpose vehicles and long-term contractual arrangements.
Nvidia has also been pursuing major financing initiatives for AI infrastructure. The company has sought to support more than $500 billion in compute financing platforms, with major financial institutions including Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR involved in various efforts.
| AI Financing Trend | Example |
|---|---|
| Chip-backed financing | Broadcom |
| Compute financing platforms | Nvidia |
| Private-credit participation | Apollo, Blackstone and others |
| Special-purpose vehicles | Broadcom AI financing |
| Long-term capacity agreements | AI labs and data-center operators |
| Infrastructure guarantees | Nvidia and other technology companies |
The growing involvement of private capital shows that AI is becoming an infrastructure-intensive industry similar in some respects to telecommunications, energy and data-center development.
Debt Creates Both Opportunity And Risk
Large-scale debt financing can accelerate AI infrastructure deployment, but it also introduces financial risk.
AI companies and infrastructure operators need sufficient long-term revenue to cover the cost of the computing capacity being financed. If AI demand grows more slowly than expected, investors could face losses and chip suppliers could face pressure.
The risk is particularly important because some AI infrastructure projects depend on long-term assumptions about model demand, cloud usage and enterprise adoption.
| Potential Benefit | Potential Risk |
|---|---|
| Faster AI infrastructure deployment | Higher leverage |
| More computing capacity | Customer default risk |
| Faster chip adoption | Demand projections may disappoint |
| Private capital mobilization | Complex financing structures |
| Long-term chip contracts | Concentration among major AI customers |
The use of special-purpose vehicles can isolate some project-level risks, but it does not eliminate them.
Blackstone And Apollo Are Becoming Major AI Financiers
Blackstone and Apollo are increasingly involved in financing the physical infrastructure behind the AI boom.
Their earlier partnership with Broadcom helped launch the $35 billion AI XPV Platform. The latest financing discussions could deepen their exposure to AI chips and data centers.
This is significant because private-equity and private-credit firms have historically focused on infrastructure, real estate and corporate lending. AI data centers are creating a new category of infrastructure assets with enormous capital requirements.
Why Private Capital Is Entering AI
Traditional banks may be reluctant to finance projects with uncertain technology and demand profiles at the scale required by AI companies.
Private-credit investors can potentially provide larger, customized financing packages and accept more complex risk structures.
This creates a complementary relationship between technology companies and financial institutions.
The Broader AI Financing Race Is Accelerating
Broadcom’s proposed financing comes only days after another major AI infrastructure financing announcement.
Nvidia has committed up to $105 billion in guarantees supporting OpenAI’s 20-year lease of a massive data center project in Ohio. The facility is expected eventually to reach 8 gigawatts of capacity, with 800 megawatts targeted to come online by 2028.
| AI Infrastructure Project | Financing/Guarantee | Capacity |
|---|---|---|
| Broadcom proposed financing | $60B+; potentially $100B | AI chip infrastructure |
| Broadcom-Apollo-Blackstone AI XPV | $35B | 1+ GW initially; 20+ GW target |
| Nvidia-OpenAI Ohio project | Up to $105B guarantee | Up to 8 GW |
| OpenAI-Broadcom custom chips | — | 10 GW |
These transactions illustrate how the AI industry is moving toward infrastructure commitments measured in tens or hundreds of billions of dollars.
What The Deal Means For Broadcom
For Broadcom, the financing strategy could strengthen its position as a critical supplier of custom AI chips and networking infrastructure.
Instead of simply selling chips to AI companies, Broadcom can potentially participate in a financing ecosystem that helps customers obtain the computing capacity needed to deploy those chips.
That creates a stronger connection between Broadcom’s future revenue and the expansion of AI infrastructure.
However, the strategy also increases exposure to a small group of extremely large customers and projects.
The Bigger Picture
Broadcom’s proposed $60 billion-plus financing deal shows how quickly the economics of AI infrastructure are changing. The potential $100 billion package would be among the largest debt structures connected to AI computing and could help finance custom chips and related infrastructure for companies such as Anthropic and potentially other major AI developers.
The transaction also demonstrates the growing role of private capital in the AI buildout. Broadcom’s earlier $35 billion AI XPV Platform with Apollo and Blackstone, Nvidia’s major infrastructure guarantees and the rapid expansion of custom AI chips all point to the same trend: AI companies increasingly need financing structures that can support infrastructure spending far beyond what operating cash flows alone can provide.
Looking Ahead
The immediate focus will be whether Broadcom can finalize the proposed financing and on what terms. The final size could be below the potential $100 billion ceiling, and the exact customers, debt structure, guarantees and investor commitments remain subject to negotiations. If completed, the transaction would give Broadcom another major mechanism for supporting AI infrastructure while potentially securing long-term demand for its custom chips and networking technology.
More broadly, the AI industry is entering a phase in which access to capital could become nearly as important as access to chips, electricity and data centers. The emergence of billion-dollar and even $100 billion-scale financing structures suggests that AI infrastructure is becoming a distinct asset class for banks and private-credit investors. The challenge will be ensuring that the enormous computing capacity being financed generates enough economic value to justify the debt being accumulated across the ecosystem.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
