{"ok":true,"count":22,"last_updated":"2026-10-26","filters":{},"summary":{"total_commitments":22,"total_amount_usd_max":718890300000,"by_silicon_dependency":{"mixed":8,"trainium":1,"tpu":2,"nvidia":11},"by_type":{"convertible-note":1,"warrant":4,"private-equity":10,"compute-commitment":5,"public-equity":2},"by_from":{"Broadcom":1,"Qualcomm":1,"Nvidia":8,"AMD":2,"Reflection AI":1,"Google":2,"Anthropic":3,"Amazon":2,"Microsoft":2}},"commitments":[{"id":"broadcom-anthropic-convertible-42b-2026-10","from":"Broadcom","to":"Anthropic","amount_usd_max":42000000000,"amount_usd_disclosed":null,"announced_date":"2026-10-01","type":"convertible-note","recipient_silicon_dependency":"mixed","commercial_quid_pro_quo":"Proceeds are restricted to Anthropic's TPU computing capacity lease obligations, so the money comes back to Broadcom as payment for hardware it co-designs and sells. Broadcom ends up as supplier and lender at the same time, a conflict the prospectus flags in its risk factors.","source_urls":["https://www.sec.gov/Archives/edgar/data/0002130357/000162828026067168/anthropicpbc-sx1.htm","https://www.cnbc.com/2026/10/01/broadcom-lending-anthropic-42-billion-chips-reuters.html"],"notes":"Disclosed in Anthropic's IPO prospectus and first reported October 1, 2026. Broadcom agreed to lend up to $42B through convertible notes that can convert into Anthropic shares; the facility would cover roughly a third of the $125.2B five-year TPU capacity lease Anthropic has committed to. No notes had been issued as of August 2, 2026, and Anthropic does not expect any to be sold before the IPO completes. No interest rate or conversion price was disclosed. The only convertible-note row here and the clearest case of vendor financing in the set."},{"id":"qualcomm-amazon-2026-09","from":"Qualcomm","to":"Amazon","amount_usd_max":60000000000,"amount_usd_disclosed":null,"announced_date":"2026-09-03","type":"warrant","recipient_silicon_dependency":"trainium","commercial_quid_pro_quo":"A multi-generation agreement for Amazon to buy Qualcomm server chip products, technology, systems and manufacturing services, reported to cover custom AI inference silicon and optical connectivity. The warrant only pays off for Amazon if Amazon keeps buying, so it converts a customer relationship into an equity interest.","source_urls":["https://www.sec.gov/Archives/edgar/data/804328/000110465926105718/tm2623289d1_8k.htm","https://aibusiness.com/data-centers/qualcomm-forges-ai-chip-deal-amazon"],"notes":"Equity flows the opposite way from most rows here: the chip vendor granted the customer a warrant. Qualcomm issued Amazon.com NV Investment Holdings LLC a warrant for up to 25 million Qualcomm shares at $161.26, expiring September 3, 2036, with 3,750,000 shares vested at issuance and the rest vesting in tranches tied to commercial arrangements, binding purchase orders and actual purchases. amount_usd_max is the $60B maximum payment level in the 8-K that would fully vest the warrant, which is a ceiling and not a committed spend. Amazon is tagged trainium because its own silicon remains its primary AI accelerator line."},{"id":"nvidia-huggingface-2026-09","from":"Nvidia","to":"Hugging Face","amount_usd_max":12930300000,"amount_usd_disclosed":null,"announced_date":"2026-09-03","type":"private-equity","recipient_silicon_dependency":"mixed","commercial_quid_pro_quo":"Ownership of the main open-model distribution hub (18M+ developers, 3M+ models, 200K+ companies) and the evaluation, inference and deployment surface on top of it. Nvidia committed in its 8-K to keep the platform open and to keep supporting other silicon vendors; its blog says Nvidia compute will not be required.","source_urls":["https://www.sec.gov/Archives/edgar/data/1045810/000104581026000078/nvda-20260902.htm","https://blogs.nvidia.com/blog/nvidia-to-acquire-hugging-face/","https://www.cnbc.com/2026/09/03/nvidia-agrees-to-buy-hugging-face-for-almost-13-billion-ai-expansion.html"],"notes":"Full acquisition, not a minority stake: definitive agreement signed September 2 for about $11.9B payable to stockholders plus an equity retention program of up to about $1.0B for staff joining Nvidia. Expected to close in the first half of 2027 subject to regulatory approvals, so nothing is drawn yet. Recipient tagged mixed because the platform serves models across clouds and accelerators."},{"id":"nvidia-ssi-2026-07","from":"Nvidia","to":"Safe Superintelligence","amount_usd_max":5000000000,"amount_usd_disclosed":null,"announced_date":"2026-07-27","type":"private-equity","recipient_silicon_dependency":"tpu","commercial_quid_pro_quo":"Access to the Vera Rubin platform that SSI says will raise its compute by an order of magnitude, plus collaboration on current and future Nvidia platforms. Moves a lab that named Google Cloud TPUs as its primary compute in 2025 onto Nvidia systems.","source_urls":["https://nvidianews.nvidia.com/news/ilya-sutskevers-safe-superintelligence-inc-and-nvidia-announce-long-term-strategic-partnership","https://www.bloomberg.com/news/articles/2026-07-27/nvidia-makes-substantial-investment-in-sutskever-s-ai-startup","https://www.datacenterdynamics.com/en/news/ai-startup-safe-superintelligence-to-use-googles-tpu-chips-for-research/"],"notes":"Nvidia's release describes a substantial equity investment without a figure; the $5B amount is Bloomberg's reporting, not a company disclosure. Silicon tag reflects SSI at announcement time (Google Cloud TPUs); the deal is structured to shift that."},{"id":"amd-anthropic-2026-07","from":"AMD","to":"Anthropic","amount_usd_max":5000000000,"amount_usd_disclosed":null,"announced_date":"2026-07-22","type":"private-equity","recipient_silicon_dependency":"mixed","commercial_quid_pro_quo":"Anthropic deploys up to 2 GW of AMD Instinct MI450-series GPUs in Helios racks (first gigawatt from the first half of 2027), building on existing MI355X use. Multi-year work using Claude to optimize Instinct workloads and speed up ROCm, and AMD adopts Claude across engineering.","source_urls":["https://ir.amd.com/news-events/press-releases/detail/1292/amd-and-anthropic-announce-strategic-partnership-to-deploy-up-to-2-gigawatts-of-amd-instinct-mi450-series-gpus","https://www.cnbc.com/2026/07/22/amd-anthropic-ai-chip-investment.html"],"notes":"AMD \"committed to make a strategic equity investment of up to $5 billion in Anthropic in the future\", so no amount is drawn yet. Adds AMD GPUs alongside the Trainium, TPU and Nvidia capacity Anthropic already contracts. Anthropic's October 26, 2026 S-1 restates it as a July hardware supply framework agreement for Helios systems under which AMD would buy up to $5.0B of Anthropic stock once conditions are met; the funds had not been received."},{"id":"reflection-spacex-compute-2026-06","from":"Reflection AI","to":"SpaceX","amount_usd_max":6300000000,"amount_usd_disclosed":null,"announced_date":"2026-06-22","type":"compute-commitment","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Immediate access to Nvidia GB300 systems at Colossus 2 for $150M a month from July 1, 2026 through 2029. Either side can exit with 90 days' notice after the first three months.","source_urls":["https://www.cnbc.com/2026/06/22/spacex-ai-colossus-data-center-reflection.html","https://www.datacenterdynamics.com/en/news/spacex-secures-63bn-compute-capacity-deal-from-ai-startup-reflection/"],"notes":"Terms come from contract materials viewed by CNBC; Reflection confirmed the agreement. The $6.3B maximum assumes the contract runs its full term. Reflection is Nvidia-backed and had not shipped a public frontier model at signing."},{"id":"google-spacex-compute-2026-06","from":"Google","to":"SpaceX","amount_usd_max":30360000000,"amount_usd_disclosed":null,"announced_date":"2026-06-05","type":"compute-commitment","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Google pays $920M a month from October 2026 through June 2029 for about 110,000 Nvidia GPUs plus CPUs and memory. Google called it short-term bridge capacity for Gemini Enterprise demand; both sides can terminate on 90 days' notice after December 31, 2026.","source_urls":["https://techcrunch.com/2026/06/05/google-will-pay-spacex-920m-per-month-for-compute/","https://aibusiness.com/generative-ai/google-pay-spacex-30-billion-ai-compute"],"notes":"Disclosed in a SpaceX regulatory filing a week before its IPO. amount_usd_max is the filing's monthly fee times the 33-month term, before the reduced ramp-up fee and any GPU-delivery shortfall credits. A TPU owner buying Nvidia capacity, so it runs against the investor-silicon loop."},{"id":"nvidia-iren-may-2026","from":"Nvidia","to":"IREN","amount_usd_max":2100000000,"amount_usd_disclosed":null,"announced_date":"2026-05-07","type":"warrant","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Intent to deploy up to 5 GW of Nvidia DSX-aligned AI infrastructure across IREN's pipeline over time, with the 2 GW Sweetwater, Texas campus as the flagship DSX site.","source_urls":["https://nvidianews.nvidia.com/news/nvidia-and-iren-announce-strategic-partnership-to-accelerate-deployment-of-up-to-5-gigawatts-of-ai-infrastructure","https://www.cnbc.com/2026/05/09/nvidia-embraces-ai-investor-topping-40-billion-in-equity-bets-2026.html"],"notes":"Structured as a five-year right to buy up to 30 million IREN shares at $70 each, subject to conditions including regulatory approval. Cleanest read of the customer-investor loop: equity option in, DSX system deployments out."},{"id":"anthropic-spacexai-colossus-2026-05","from":"Anthropic","to":"SpaceX","amount_usd_max":45000000000,"amount_usd_disclosed":null,"announced_date":"2026-05-06","type":"compute-commitment","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Anthropic rents all available compute at Colossus 1 near Memphis, later extended to Colossus 2 (Anthropic's Series H post cites access to both). Stated interest in multi-gigawatt orbital compute capacity with SpaceX.","source_urls":["https://www.sec.gov/Archives/edgar/data/1181412/000162828026036936/spaceexplorationtechnologi.htm","https://www.cnbc.com/2026/05/06/anthropic-spacex-data-center-capacity.html","https://www.datacenterdynamics.com/en/news/spacex-ipo-filing-reveals-anthropic-set-to-pay-musks-firm-125bn-a-month-to-rent-xai-data-center-space/","https://www.anthropic.com/news/series-h"],"notes":"Announced May 6 with no terms. SpaceX's S-1 (filed May 20) disclosed $1.25B a month through May 2029, terminable on 90 days' notice; amount_usd_max is that rate over roughly 36 months. xAI, which built Colossus, is part of SpaceX (SpaceXAI)."},{"id":"nvidia-corning-may-2026","from":"Nvidia","to":"Corning","amount_usd_max":3200000000,"amount_usd_disclosed":500000000,"announced_date":"2026-05-06","type":"warrant","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Three new US optical connectivity plants in North Carolina and Texas, a 10x expansion of Corning's US optical connectivity capacity and 50%+ more US fiber capacity, aimed at Nvidia-scale AI factory buildouts.","source_urls":["https://www.corning.com/worldwide/en/about-us/news-events/news-releases/2026/05/nvidia-and-corning-announce-long-term-partnership-to-strengthen-us-manufacturing-for-ai-infrastructure.html","https://www.cnbc.com/2026/05/06/nvidia-corning-optical-factories-nc-texas-ai.html"],"notes":"Per CNBC: a $500M pre-funded warrant for up to 3 million Corning shares plus warrants for up to 15 million more at $180. Capital is the lever to commit Corning to fab capex on Nvidia's timeline rather than acquire Corning outright."},{"id":"anthropic-google-tpu-200b-2026","from":"Anthropic","to":"Google Cloud + Broadcom","amount_usd_max":200000000000,"amount_usd_disclosed":null,"announced_date":"2026-05-05","type":"compute-commitment","recipient_silicon_dependency":"tpu","commercial_quid_pro_quo":"Five gigawatts of next-generation TPU capacity with Google and Broadcom, coming online from 2027. Google effectively recollects much of its up-to-$40B Anthropic equity commitment on the compute side.","source_urls":["https://www.theinformation.com/articles/anthropic-200-billion-google-cloud-tpu-deal","https://www.anthropic.com/news/series-h"],"notes":"Direction reversed: Anthropic committing spend to Google Cloud, not equity flowing to Anthropic. The 5 GW capacity deal was announced in April and restated in the May 28 Series H post; the roughly $200B five-year value comes from The Information (May 5) and has not been confirmed by either company. Anthropic's October 26, 2026 S-1 splits the relationship into at least $114.5B committed to Google through July 2033 under a March 2026 agreement, payable regardless of actual usage, plus a $125.2B five-year lease of TPU capacity on Broadcom-built hardware. Those figures overlap this row rather than adding to it, so amount_usd_max is left at the reported deal value."},{"id":"google-anthropic-equity-40b","from":"Google","to":"Anthropic","amount_usd_max":40000000000,"amount_usd_disclosed":10000000000,"announced_date":"2026-04-24","type":"private-equity","recipient_silicon_dependency":"mixed","commercial_quid_pro_quo":"Deepens the TPU relationship: Anthropic had secured 5 GW of Google and Broadcom TPU capacity earlier in April. Anthropic runs production workloads on Trainium, TPU and Nvidia silicon.","source_urls":["https://www.cnbc.com/2026/04/24/google-to-invest-up-to-40-billion-in-anthropic-as-search-giant-spreads-its-ai-bets.html","https://www.bloomberg.com/news/articles/2026-04-24/google-plans-to-invest-up-to-40-billion-in-anthropic"],"notes":"Google put in $10B of cash at the Series G price ($350B pre-money) with up to $30B more contingent on performance milestones. Before this deal Google had invested a little over $3B for a reported 14% stake. Pairs with the TPU compute commitment as one half of a circular flow."},{"id":"amazon-anthropic-25b-2026-04","from":"Amazon","to":"Anthropic","amount_usd_max":25000000000,"amount_usd_disclosed":5000000000,"announced_date":"2026-04-20","type":"private-equity","recipient_silicon_dependency":"mixed","commercial_quid_pro_quo":"Anthropic commits more than $100B of AWS spend over ten years for up to 5 GW of capacity across Trainium2, Trainium3 and Trainium4, plus tens of millions of Graviton cores. AWS stays Anthropic's primary cloud and training partner.","source_urls":["https://www.aboutamazon.com/news/company-news/amazon-invests-additional-5-billion-anthropic-ai","https://www.cnbc.com/2026/04/20/amazon-invest-up-to-25-billion-in-anthropic-part-of-ai-infrastructure.html","https://www.anthropic.com/news/series-h"],"notes":"$5B invested at signing (Anthropic's Series H counts it as previously committed hyperscaler money) and up to $20B more tied to commercial milestones. On top of the $8B Amazon invested from 2023, for a total commitment of up to $33B. Anthropic's October 26, 2026 S-1 puts the matching spend commitment at no less than $110B to Amazon through 2036."},{"id":"nvidia-nebius-2026-03","from":"Nvidia","to":"Nebius","amount_usd_max":2000000000,"amount_usd_disclosed":null,"announced_date":"2026-03-11","type":"public-equity","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Enables Nebius to deploy more than 5 GW of capacity by the end of 2030, with Nvidia supporting early adoption of the newest generation of its accelerated computing platform.","source_urls":["https://www.sec.gov/Archives/edgar/data/1513845/000110465926026163/tm268532d1_ex99-1.htm","https://www.cnbc.com/2026/03/11/nebius-nvidia-ai-cloud.html"],"notes":"Joint release filed as exhibit 99.1 to a Nebius Form 6-K. It states the $2B figure and the 5 GW target but names no instrument, valuation or equity terms, so nothing is recorded as drawn. Nvidia was already a Nebius shareholder from the December 2024 round. Backfilled in the October 2026 refresh."},{"id":"amazon-openai-50b-2026-02","from":"Amazon","to":"OpenAI","amount_usd_max":50000000000,"amount_usd_disclosed":50000000000,"announced_date":"2026-02-27","type":"private-equity","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"OpenAI expands its $38B AWS agreement by $100B over eight years, including about 2 GW of Trainium3 and Trainium4 capacity. AWS becomes the exclusive third-party cloud distributor for OpenAI Frontier, and the two co-develop a Stateful Runtime Environment on Bedrock.","source_urls":["https://www.aboutamazon.com/news/aws/amazon-open-ai-strategic-partnership-investment","https://www.cnbc.com/2026/02/27/open-ai-funding-round-amazon.html","https://www.pymnts.com/news/artificial-intelligence/2026/amazon-completes-50-billion-dollar-investment-openai/"],"notes":"Announced as $15B up front and $35B more when conditions were met. Amazon's July 31, 2026 SEC filing said it had invested all $50B ($15B in Q1, $13.7B in Q2, $21.3B after June 30). Part of the OpenAI round that closed March 31 at $852B post-money."},{"id":"nvidia-openai-feb-2026","from":"Nvidia","to":"OpenAI","amount_usd_max":30000000000,"amount_usd_disclosed":null,"announced_date":"2026-02-27","type":"private-equity","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Multi-year compute orders and roadmap alignment on the silicon side; OpenAI's GPT-5.5 was co-designed for, trained with and served on Nvidia GB200 and GB300 NVL72 systems. Functions as a long-dated option on OpenAI staying primarily on Nvidia silicon.","source_urls":["https://www.cnbc.com/2026/02/27/open-ai-funding-round-amazon.html","https://openai.com/index/accelerating-the-next-phase-ai/","https://www.cnbc.com/2026/05/09/nvidia-embraces-ai-investor-topping-40-billion-in-equity-bets-2026.html"],"notes":"Nvidia's $30B in the OpenAI round announced at $110B on February 27 and closed at $122B of committed capital on March 31, at an $852B post-money valuation. Reporting at the time said the $30B replaced the September 2025 letter of intent under which Nvidia would have invested up to $100B, which is why that larger figure has no row here. CNBC called it Nvidia's single largest bet among more than $40B of 2026 equity commitments as of May."},{"id":"amd-meta-2026-02","from":"AMD","to":"Meta","amount_usd_max":100000000000,"amount_usd_disclosed":null,"announced_date":"2026-02-24","type":"warrant","recipient_silicon_dependency":"mixed","commercial_quid_pro_quo":"Meta deploys up to 6 GW of AMD Instinct GPUs across multiple generations, starting with a custom MI450 based GPU paired with 6th Gen EPYC Venice CPUs in AMD Helios racks that the two companies developed through the Open Compute Project. Shipments for the first gigawatt begin in the second half of 2026, and the roadmaps for GPU, CPU, systems and software are aligned.","source_urls":["https://ir.amd.com/news-events/press-releases/detail/1279/amd-and-meta-announce-expanded-strategic-partnership","https://www.tomshardware.com/tech-industry/artificial-intelligence/amd-meta-100-billion-deal"],"notes":"Same inverted structure as the Qualcomm and Amazon row: AMD issued Meta a performance based warrant for up to 160 million AMD shares, with the first tranche vesting on the initial 1 GW of shipments, further tranches vesting as purchases scale to 6 GW, vesting also conditioned on AMD share price thresholds and exercise conditioned on Meta hitting technical and commercial milestones. AMD disclosed no dollar value; amount_usd_max is the more than $100B deal value reported by the Wall Street Journal and is not a company figure. Meta is tagged mixed: it runs Nvidia at scale plus its own MTIA silicon and already had MI300 and MI350 series deployments."},{"id":"nvidia-coreweave-2026-01","from":"Nvidia","to":"CoreWeave","amount_usd_max":2000000000,"amount_usd_disclosed":2000000000,"announced_date":"2026-01-26","type":"public-equity","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Buildout of more than 5 GW of AI factories by 2030 on Nvidia's accelerated computing platform, with early adoption of coming generations including Rubin, Vera CPUs and BlueField storage. Nvidia also puts its balance sheet behind CoreWeave's hunt for land, power and shell space, and will test CoreWeave's SUNK and Mission Control software for possible inclusion in Nvidia reference architectures.","source_urls":["https://www.coreweave.com/news/nvidia-and-coreweave-strengthen-collaboration-to-accelerate-buildout-of-ai-factories","https://www.cnbc.com/2026/01/26/3coreweave-nvidia-stock-ai-data-centers.html"],"notes":"Nvidia invested $2B in CoreWeave Class A common stock at $87.20 a share, on top of the roughly 7% stake it disclosed before CoreWeave's 2025 IPO. Sits alongside the September 2025 agreement under which Nvidia buys unsold CoreWeave capacity for $6.3B, making Nvidia investor, supplier and customer at once. Backfilled in the October 2026 refresh."},{"id":"anthropic-azure-30b-2025-11","from":"Anthropic","to":"Microsoft Azure","amount_usd_max":30000000000,"amount_usd_disclosed":null,"announced_date":"2025-11-18","type":"compute-commitment","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Anthropic buys $30B of Azure compute and may contract up to 1 GW more, starting on Nvidia Grace Blackwell and Vera Rubin systems. Claude comes to Microsoft Foundry and stays in the Copilot family.","source_urls":["https://www.anthropic.com/news/microsoft-nvidia-anthropic-announce-strategic-partnerships","https://blogs.microsoft.com/blog/2025/11/18/microsoft-nvidia-and-anthropic-announce-strategic-partnerships/"],"notes":"The compute half of the November 2025 three-way deal in which Microsoft and Nvidia committed equity to Anthropic. Anthropic's October 26, 2026 S-1 shows the agreement was later amended: it now commits at least $44.9B to Microsoft through May 2033, regardless of actual usage. amount_usd_max stays at the $30B announced in November 2025 because that is the figure attached to this announcement. Backfilled in the September 2026 refresh."},{"id":"nvidia-anthropic-10b-2025-11","from":"Nvidia","to":"Anthropic","amount_usd_max":10000000000,"amount_usd_disclosed":null,"announced_date":"2025-11-18","type":"private-equity","recipient_silicon_dependency":"mixed","commercial_quid_pro_quo":"First deep technology partnership between the two: joint design work to optimize Claude for Nvidia systems and future Nvidia architectures for Anthropic workloads, with Anthropic's initial commitment up to 1 GW of Grace Blackwell and Vera Rubin compute.","source_urls":["https://www.anthropic.com/news/microsoft-nvidia-anthropic-announce-strategic-partnerships","https://www.cnbc.com/2025/11/18/anthropic-ai-azure-microsoft-nvidia.html"],"notes":"Commitment of up to $10B. Anthropic's February 2026 Series G post says the round included a portion of the previously announced Microsoft and Nvidia investments; the drawn amount was not disclosed. Backfilled in the September 2026 refresh."},{"id":"microsoft-anthropic-5b-2025-11","from":"Microsoft","to":"Anthropic","amount_usd_max":5000000000,"amount_usd_disclosed":null,"announced_date":"2025-11-18","type":"private-equity","recipient_silicon_dependency":"mixed","commercial_quid_pro_quo":"Paired with Anthropic's $30B Azure compute commitment; puts Claude on Azure and Microsoft Foundry and keeps it inside GitHub Copilot, Microsoft 365 Copilot and Copilot Studio.","source_urls":["https://blogs.microsoft.com/blog/2025/11/18/microsoft-nvidia-and-anthropic-announce-strategic-partnerships/","https://www.cnbc.com/2025/11/18/anthropic-ai-azure-microsoft-nvidia.html"],"notes":"Commitment of up to $5B. A portion was folded into the February 2026 Series G; the drawn amount was not disclosed. Backfilled in the September 2026 refresh."},{"id":"microsoft-openai-2019-2023","from":"Microsoft","to":"OpenAI","amount_usd_max":13000000000,"amount_usd_disclosed":13000000000,"announced_date":"2023-01-23","type":"private-equity","recipient_silicon_dependency":"nvidia","commercial_quid_pro_quo":"Exclusive Azure cloud partnership for OpenAI training and inference, profit-share rights, deep API integration into Microsoft products.","source_urls":["https://blogs.microsoft.com/blog/2023/01/23/microsoftandopenaiextendpartnership/"],"notes":"Multi-tranche through 2019-2023 totaling roughly $13B. The original anchor of the modern AI capital stack; Microsoft added about $750M in OpenAI's October 2024 round."}],"attribution":{"curator":"TensorFeed.ai editorial","curator_url":"https://tensorfeed.ai/funding","underlying_sources":"SEC filings, hyperscaler press releases, reputable trade reporting (CNBC, The Information, etc.). Each entry carries source_urls citing the public record we drew from.","license":"TF curation, tagging, and commercial_quid_pro_quo summaries are CC-BY 4.0. Underlying public-record sources retain their original license.","notes":"Hand-curated. Update cadence is editorial: an entry is added when a related original article goes out, or when reporting picks up something we have not covered."}}