1. What is the SoftBank AI fund being discussed?
SoftBank Group chief executive Masayoshi Son is seeking to raise as much as $100 billion from Gulf investors for a new SoftBank AI fund, according to people familiar with the discussions, with the talks reported on October 9, 2026. Son has held preliminary conversations in recent weeks with senior figures, including in the United Arab Emirates. SoftBank has not confirmed the plan, and no investors have been named.
The proposed vehicle would not resemble the venture funds that made SoftBank famous. Rather than writing checks to young companies developing frontier artificial intelligence, the fund would buy established businesses and then upgrade them with AI and other advanced technologies. In other words, the target is the large part of the economy that does not yet run on the newest tools, and the thesis is that applying existing AI to companies already earning revenue can create value without the risk of inventing the technology first.
The scale is the first thing that stands out. A $100 billion raise would put the fund among the largest private pools of capital ever assembled for a single theme, and it would arrive only days after SoftBank had been in the headlines for its heavy commitments to the AI sector. The talks are at an early stage, and people close to them cautioned that there is no guarantee they will succeed.
2. How would the fund put AI to work inside older companies?
The plan reportedly leans on Roze, SoftBank's robotics and physical-AI business, which is expected to play a central role. The idea is to acquire companies outside the technology sector and then use Roze's robotics and AI systems to improve how they operate. That could mean automating physical processes, trimming costs, or raising output in industries such as logistics or manufacturing, although the reported plan does not name specific companies, sectors or targets.
This approach resembles a familiar private equity playbook, in which an investor buys a business, changes how it is run and aims to sell it later at a higher value. The difference is the lever being pulled. Instead of financial engineering or management changes alone, the fund would rely on a technology platform that SoftBank itself owns. That ties the fund closely to SoftBank's in-house capabilities and raises the stakes on whether Roze can deliver measurable gains in settings far removed from a software laboratory.
For prospective investors the appeal is straightforward. A portfolio of operating businesses generates cash flow, unlike a bet on a startup that may take years to earn a profit. The risk is execution: integrating robotics into existing factories, warehouses or service networks is slow and expensive, and results depend on the quality of each target and of the management teams that run them.
3. Why is SoftBank turning to Gulf investors?
Gulf capital is familiar territory for Son. When he launched the first Vision Fund in 2017, which raised nearly $100 billion, the money came largely from sovereign wealth in the region, most notably Saudi Arabia's Public Investment Fund and Abu Dhabi's Mubadala. Returning to the same pool of investors for a new mega-fund follows a pattern in which SoftBank pairs its deal-making reputation with the deep reserves of state-backed investors.
Gulf states have also been vocal about their ambitions in artificial intelligence, and the UAE in particular has positioned itself as a hub for computing infrastructure and AI investment. A fund focused on applying AI across established industries gives those investors exposure to the technology without limiting them to a handful of model developers. It also offers a way to diversify national economies that have historically depended on energy revenue.
The original Vision Fund left a mixed legacy. It produced some high-profile gains and sizable losses, and it remains one of the most debated investment vehicles of the past decade. Any new approach to the same investors will invite comparison with that record, and prospective backers are likely to ask how the new strategy differs from the earlier one and what safeguards would apply to valuations and concentration.
4. How much has SoftBank already committed to AI?
The fundraising effort follows a heavy period of spending. SoftBank has made a $30 billion follow-on investment in OpenAI, the developer of ChatGPT, and the company raised $11.1 billion through a bond sale in September 2026. Those commitments show how much of the group's balance sheet is already tied to the fortunes of a small number of AI companies.
Critics have pointed to several pressure points. One is concentration risk, since a large share of SoftBank's AI exposure sits in a single company. Another is timing: the payoff from a possible OpenAI public listing may be some way off, leaving SoftBank to carry its commitments in the meantime. A third is the rising debt load that has come with funding these commitments through the bond market.
Seen in that light, a Gulf-backed fund has a second attraction beyond the investment case itself. Outside capital would allow SoftBank to pursue a large new strategy without placing the whole cost on its own balance sheet, while earning management income from the vehicle. Whether investors see it that way will depend on the terms offered, which have not been disclosed.
5. Why does a $100 billion figure matter for the wider market?
Fundraising of this size can influence more than SoftBank's own fortunes. If a pool of capital that large were assembled for buying and modernizing operating businesses, it could raise competition for acquisition targets, lift prices for companies seen as good candidates for automation, and encourage rival investors to build comparable strategies.
It would also test appetite among sovereign investors for AI-linked assets outside the best-known developers. Their willingness to back an acquisition-led model rather than a venture-led one would indicate how they judge the balance between the potential and the cost of the technology. For companies that might become targets, the prospect of a well-funded buyer promising robotics and AI upgrades could be attractive, or unsettling, depending on how it affects jobs and ownership. None of this is certain, because the plan remains at an early stage, but the scale alone makes it a development that market participants are likely to follow closely.
6. What does the shift suggest about the stage of the AI cycle?
The change in focus carries a message about where the technology cycle stands. Early in a major technology wave, capital tends to flow to the inventors: the laboratories and startups building the underlying tools. Later, money starts to look for ways to profit from the technology across the wider economy, by applying it to businesses that already have customers, assets and revenue.
A fund that buys established companies instead of backing new model builders fits that second phase. It implies that at least some large investors believe the core technology is mature enough to be deployed at scale, and that the next gains may come from implementation rather than invention. That reading would also help explain the characterization of the push as a sign of a late-stage boom.
There is a cautionary side to the same observation. Late-stage phases of any investment cycle are often the period when valuations are stretched and enthusiasm is highest. A shift toward buying existing businesses can reflect genuine maturity, but it can also signal that the most attractive early opportunities have already been priced in. The article of faith behind the fund, that AI can reliably improve the economics of ordinary companies, has yet to be proven at the scale SoftBank is contemplating.
7. What are the risks and open questions?
Several uncertainties hang over the plan. SoftBank has declined to comment on the talks, and the reported figure of up to $100 billion describes an ambition rather than a commitment. Talks of this kind often change in size, structure and participants before any deal is announced, and many never reach a final agreement.
There are also operational questions. Buying established companies at scale requires a pipeline of suitable targets, careful due diligence and the ability to manage a diverse set of businesses at once. Applying robotics and AI to them adds a technology-delivery challenge on top of ordinary ownership duties. The reported role for Roze concentrates much of that burden on one part of the SoftBank group.
Finally, there is the matter of governance and concentration. Investors who remember the first Vision Fund will want clarity about decision-making, the limits on any single position and how conflicts between SoftBank's own holdings and the fund's investments would be handled. The details of those protections would likely shape how much capital Gulf investors are willing to commit.
8. What to watch next
The first signal will be whether SoftBank confirms the plan publicly, since the company has so far declined to discuss the talks. Confirmation would likely come with details on the fund's size, structure and investment mandate, and on how it relates to SoftBank's existing vehicles.
The second is which Gulf investors, if any, make firm commitments, and how much of the reported $100 billion is actually raised. A smaller first close or a longer fundraising period would indicate that investors are weighing the strategy carefully.
A third marker is the fate of SoftBank's existing AI holdings. A public listing of OpenAI could convert SoftBank's paper stake into liquid value and ease the pressure from its debt-funded spending, which in turn would shape how aggressively the group can pursue new strategies. Until those pieces fall into place, the $100 billion fund remains a proposal that shows where one of the technology sector's most influential investors thinks the next phase of the AI cycle is heading.