SoftBank Plans to Raise Up to $100 Billion from Gulf Investors for AI Investments
Sources say SoftBank founder Masayoshi Son is seeking to raise up to $100 billion from Gulf investors to boost large-scale artificial intelligence investments. In recent weeks, he has consulted with several high-level executives, including individuals from the UAE. This matter has not been confirmed by the company, and external verification has not been possible.
The funds are intended to establish a fund to acquire other companies and then use AI and other advanced technologies to transform the operations of the acquired companies. The related concept also mentions robotics technology, with the approach being to first acquire companies and then change their production and management methods, rather than just buying equity in model companies.
The timing closely follows SoftBank's existing bets. Last week, the company announced it had completed a $30 billion investment in OpenAI, corresponding to its committed share in the last round of financing for ChatGPT's parent company. A month earlier, SoftBank issued $11.1 billion in high-yield corporate bonds, the largest of its kind globally at that time, to fund this massive bet on OpenAI.
If the new fund materializes, the funding structure will shift from primarily using its own balance sheet and issuing debt to managing an external capital pool. Gulf investors will provide most of the cash, while SoftBank will be responsible for identifying acquisition targets and integrating AI and robotics transformations. Currently, negotiations are still in the early stages, and the amount is a ceiling, not a confirmed commitment.
The buyers are Gulf region investors, and the sellers are companies willing to be acquired and accept operational transformations. The transactions are driven by AI capital expenditures, with funds flowing from sovereign and institutional capital into the acquisition fund, and then into the invested companies. If SoftBank gains management rights, it can obtain project sources and transformation profits; the OpenAI-related ecosystem and targets that can be transformed by AI will benefit; while SoftBank's own balance sheet and investors holding its high-yield bonds will continue to bear the financing and valuation pressures from concentrated bets.
Source: Public Information
ABAB AI Insight
Masayoshi Son's last major Gulf funding initiative was the Vision Fund, which closed in 2017 with approximately $45 billion from the Saudi Public Investment Fund and about $15 billion from Abu Dhabi's Mubadala, totaling around $100 billion. The subsequent strategy focused on concentrated bets on Uber, WeWork, DoorDash, and ARM: ARM went public in 2023, becoming an exit asset, while WeWork filed for bankruptcy in 2023. The Vision Fund recorded significant losses over several quarters, and SoftBank's stock price fell in 2021-2022 alongside the fund's valuation decline. In 2025, to fund OpenAI, SoftBank sold approximately $5.8 billion of its Nvidia shares, reinvesting the realized chip profits back into model companies.
This time, the capital path shifts the funding source from SoftBank's balance sheet to an external fund. In December 2024, Son promised to invest $100 billion over four years in the U.S. and promoted a data center project named Stargate with OpenAI and Oracle, with an initial target of about $500 billion. By 2026, SoftBank had used $11.1 billion in high-yield bonds to cover the $30 billion investment in OpenAI and is now seeking a new $100 billion acquisition fund from the Gulf, effectively extending the same chain with sovereign funds: model equity, computing power, and then buying traditional companies for AI transformation. The Gulf side already has established channels, with Abu Dhabi's MGX receiving $49 billion in the first phase of its fund in 2026 and participating in financing for OpenAI, Anthropic, and xAI, targeting an asset scale of $100 billion.
The comparable model is not a single venture investment but rather large acquisitions made with externally committed funds before Blackstone's IPO in 2007, and 3G Capital's acquisitions of Burger King and Tim Hortons, which improved profit margins through cost restructuring. The difference is that 3G relied on layoffs and zero-based budgeting, while Son's leverage is AI and robotics. The industry position has moved past the first phase of "buying model equity" into an expansion phase using external funds to control application-layer companies; SoftBank is transitioning from a primary investor to a manager, while Gulf funds are shifting from limited partners to incremental ammunition.
Structurally, this represents capital concentration. The fixed costs of model training and inference have become too high for a single company's balance sheet to sustain. OpenAI's financing discussions have reached up to $100 billion, and SoftBank can only issue debt first and then seek sovereign investors. Whoever controls callable commitments of $100 billion holds the pricing power to acquire traditional companies and rewrite their processes; funding is dispersed in the Gulf, while decision-making is centralized at SoftBank. In case of failure, losses are borne by limited partners, while in success, the manager and a few controlling parties reap the equity gains from the transformation.