What did China decide about SoftBank's ABB Robotics deal?
China has prohibited SoftBank's purchase of ABB's robotics business in the country, ABB said on November 5, 2026. The Swiss engineering group said the decision came from the Office of the Working Mechanism for Foreign Investment Security Review in China. The office sits inside the National Development and Reform Commission, the country's top economic planner.
The deal was signed on October 8, 2025. SoftBank agreed to buy 100% of a holding company that would contain ABB Robotics. The enterprise value was $5.375 billion, or about 818.7 billion yen. The companies targeted a close in mid-to-late 2026, subject to approvals in the European Union, China and the United States.
The Financial Times first reported the prohibition, citing people familiar with the matter. Reuters, The New York Times, The Wall Street Journal and Nikkei Asia followed with their own accounts. None of the reports included an official Chinese explanation for the decision.
ABB confirmed the block in a short statement. The company said it had been informed by SoftBank, which manages the regulatory approval process. "ABB and SoftBank are in close dialogue and are working constructively on the transaction," ABB said. The wording stopped short of saying the deal had been abandoned.
The transaction has not been formally terminated. SoftBank is trying to establish what went wrong and whether it can apply again, according to the Financial Times. Neither company has announced a revised structure, a new timetable or a termination fee.
SoftBank shares fell about 5% in Tokyo morning trade on November 5. The Nikkei 225 was down 0.5%, according to InvestingLive's account of the FT report. The stock was already under pressure from Arm's quarterly guidance and a wider selloff in AI-linked shares.
Why did China block the SoftBank-ABB Robotics deal?
China has not explained the decision. People familiar with the review told the Financial Times that concerns included robotics data being transferred overseas. That data could then be used to train AI models, the people said. The Wall Street Journal reported that Beijing acted to protect technology and data from leaking to foreign rivals. Those accounts are reporting, not an official finding.
The explanation fits the economics of modern robotics. Modern industrial robots can carry cameras, force sensors and software that record how machines move, where they fail and how production lines behave. Those records are increasingly treated as training material for robot foundation models, which aim to generalize across factories and tasks.
Data is also hard to value in a diligence process. A factory robot's telemetry can reveal production volumes, defect rates and supplier layouts, which manufacturers treat as confidential. A foreign owner with access to that stream could learn more about Chinese factories than any customer contract discloses. That makes the data question a commercial issue as well as a security one.
For Beijing, letting a Japanese conglomerate control those data flows could hand a foreign rival an asset built on Chinese factory floors. The review is not a standard antitrust test. It asks whether a deal affects national security, including control over important technology and data.
SoftBank is not a passive buyer. It owns Arm and has invested heavily in OpenAI, so it could combine ABB's factory data with serious model-building capacity. That combination is exactly the kind of transfer China's review is designed to examine.
The legal basis is a set of measures that took effect on January 18, 2021. They created a cross-ministry working mechanism, with an office at the NDRC, to screen foreign investment in sensitive sectors. The list includes important information technology and internet products, important equipment manufacturing and key technologies.
The process gives the office broad discretion. It can approve a deal, attach conditions or prohibit it. Decisions are not published with detailed reasoning, and companies have little public recourse once the office rules. That opacity makes the risk hard to price before signing.
Data handling is part of the security question, not a separate compliance box. Connected robots can collect factory images, location data and operating records. Those flows cross borders when a foreign owner manages the fleet. It is not confirmed which consideration drove the ABB decision, so the data explanation remains an attribution to people familiar with the review.
The timing also matters. The review landed after the deal had already cleared Europe and while US approval was still pending. A late-stage veto is more damaging than an early one. By then, the buyer has arranged financing, carved out the business and told investors to expect a close.
Why does China have a veto over a Swiss-Japanese deal?
Neither ABB nor SoftBank is Chinese. China's leverage comes from the target's operations, not the buyer's passport. ABB Robotics sells to Chinese manufacturers, builds robots in Shanghai and employs thousands of people in the country.
China is the largest market for ABB's robotics unit, according to the Financial Times. The company's Shanghai site includes a 67,000-square-metre robot factory that opened in December 2022 after a $150 million investment. ABB described the plant as a robotics mega factory with an open lab for customer collaboration.
About 90% of the robots ABB builds for Chinese customers are manufactured locally, The Robot Report reported. That local footprint is what brings the transaction inside China's security review, even though the buyer and seller are foreign. The review follows the assets, not the headquarters.
The market is enormous. Factories in China installed 354,000 industrial robots in 2025, according to the International Federation of Robotics. That was up 20% year on year and equal to 59% of global installations. Chinese suppliers sold 195,000 units at home, a 55% share of the domestic market.
Globally, the operational stock reached 5 million industrial robots in 2025, a 9% increase, and installations exceeded 600,000 units. The United States installed about 38,500 robots, making it the second-largest market after China. Japan installed 36,219 units, a 19% decline.
A prohibition that covers the China business therefore hits the core of the asset. It is not a peripheral subsidiary. It is the largest customer base and a major manufacturing site. It is also a source of the operating data that makes the unit valuable to a physical-AI buyer.
The structure of the deal compounds the problem. SoftBank planned to buy 100% of a carved-out holding company, so control of the whole unit would pass to Tokyo. China's review can block that transfer inside its borders, which undermines the global transaction even if the parties keep talking.
ABB Group itself is far larger than the robotics unit. It employs about 110,000 people and traces its history back more than 140 years, according to the company. Robotics is a small slice of group revenue, but it is the slice most exposed to China's industrial base.
What does the block mean for SoftBank's physical AI plan?
Masayoshi Son has made physical AI a central part of SoftBank's story. When the deal was announced, he said: "SoftBank's next frontier is Physical AI." He described a plan to combine artificial superintelligence with robots. The business would be built around ABB's technology and staff.
ABB Robotics offered a shortcut. The unit brought about 7,000 employees, factories in China and the United States, and long-standing relationships with manufacturers. SoftBank could pair that hardware and customer base with its chip assets, including Arm, and its large stake in OpenAI.
Losing the unit does not end the strategy. SoftBank can invest in other robotics companies, license models or build around Arm. But the deal was the clearest route to owning a Western industrial robotics platform at scale. Minority stakes in startups are a weaker substitute.
The timing is awkward. SoftBank's shares were already falling on Arm's guidance and the AI selloff that followed the US midterm elections. A regulatory veto in China adds a political risk that investors had not priced into the robotics plan.
The Financial Times reported that SoftBank is assessing whether a fresh application is viable. That leaves the deal in limbo rather than dead. A reapplication would probably require changes that address China's concerns, and no such plan has been disclosed.
SoftBank's balance sheet gives it alternatives, but not equivalents. It can fund internal research or back startups, yet ABB's factories, service engineers and installed base would take years to rebuild. The group would also lose the customer relationships that make industrial robotics sticky.
There is also a strategic cost. Physical AI depends on real-world data from factories, warehouses and laboratories. ABB's installed machines and Chinese customer base would have supplied both hardware and data. Without them, SoftBank must build or buy those inputs elsewhere, likely at a higher price and slower pace.
What happens to ABB Robotics now?
ABB had already decided to leave robotics before SoftBank appeared. On April 17, 2025, it announced plans for a 100% spin-off. The unit would trade as a separate listed company in the second quarter of 2026. Six months later, it agreed to sell the business instead.
The unit is meaningful but not huge for ABB. It generated $2.279 billion in revenue in 2024, about 7% of group sales. EBITDA was $313 million and operational EBITA was $277 million. Revenue was $2.452 billion in 2023 and $2.258 billion in 2022, showing how cyclical the business can be.
ABB's logic for exiting was portfolio focus. The company describes itself as a leader in electrification and automation, and robotics is a cyclical division. A sale to SoftBank promised a cleaner exit than a listing, with a strategic buyer and a defined price.
ABB's third-quarter 2026 report, published on October 20, still expected the sale to close in the fourth quarter of 2026. It reported the unit as discontinued operations and pointed to an expected $2.4 billion pre-tax gain from the divestment. The group also disclosed stranded costs linked to the separation.
That gain and the clean exit are now at risk. ABB has several options, none of them easy. It could wait for SoftBank to reapply or negotiate a structure that excludes the China business. It could also revive the spin-off or keep the unit and run it for cash.
ABB has not chosen publicly. Its statement stressed dialogue and constructive work, language that suggests the parties are still looking for a route through. The company has not set a new deadline or withdrawn its financial guidance for the separation.
For ABB shareholders, the difference matters. A completed sale delivers cash and a one-off gain. A failed deal leaves a lower-margin, cyclical unit inside a group that has spent two years simplifying its portfolio. A spin-off would return the unit to public markets, but at a valuation set by investors rather than a negotiated price.
Is robotics splitting into rival blocs?
The ABB decision is part of a wider pattern. In April 2026, China's NDRC prohibited Meta's roughly $2 billion acquisition of Manus. The AI startup was founded in China and later based in Singapore. Regulators ordered the parties to unwind the deal.
The NDRC said it had decided to "prohibit foreign investment in the Manus project in accordance with laws and regulations." Morgan Lewis described it as China's first AI security review block. The message was that Chinese-origin technology and data remain under Beijing's control after a foreign sale.
The United States is building its own barriers. On July 27, 2026, the FCC determined that foreign-produced advanced robotic devices pose an unacceptable risk. It added them to its Covered List. New models cannot obtain equipment authorization; devices already authorized are exempt.
The covered category includes mobile robots such as humanoids and quadrupeds. Separately, President Trump signed a Section 232 proclamation on October 28, 2026, covering imports of robotics and industrial machinery. The proclamation imposed no immediate tariffs and directed 180 days of negotiations with trading partners.
Europe has moved the other way. The European Commission received notification of the concentration on February 13, 2026. It cleared SoftBank's acquisition of ABB Robotics without conditions on March 11, 2026, under case M.12248. Brussels treated the deal as an ordinary merger review, not a security question.
The result is three different regimes. Brussels reviews competition, Washington restricts imports and Beijing controls outbound ownership of technology and data. For robotics companies, local production, local data storage and separate supply chains are becoming conditions of doing business in the largest markets.
For Beijing, the review is also industrial policy. Chinese robot makers already lead at home, and slowing a foreign buyer's access to Chinese factories protects that position. For Washington, import restrictions serve a similar goal by raising the cost of foreign hardware.
The split has consequences beyond one deal. Companies will draft contracts that allocate regulatory risk, shorten long-stop dates and build China carve-outs into their plans. Buyers will pay less for targets with large Chinese operations, because the probability of a late veto is now visible.
What should you watch next?
SoftBank's next move is the key variable. A new application, a restructured deal or a withdrawal would each set a different path for ABB and for the unit's employees. The company has said little beyond ABB's account of the dialogue.
Watch whether China explains its reasoning. A published rationale would clarify whether the objection is data, manufacturing capacity or the transfer of dual-use technology. Silence keeps the risk premium on other deals with China exposure.
Watch the US approval process. The deal still needed American clearance as of late October, according to the Financial Times account carried by InvestingLive. Even a Chinese reversal would leave that hurdle in place.
Watch ABB's spin-off plans. If the sale collapses, a listing is the most obvious fallback. The company had already prepared for one, and investors are likely to ask whether it can be revived on a shorter timetable.
Watch China's own robotics champions. Domestic suppliers now hold 55% of installations at home, and Beijing's review gives them a protected base. That strengthens their position in the global market for industrial and humanoid robots.
Finally, watch how other cross-border robotics deals are screened. The ABB case shows that a target's China operations can decide the fate of a transaction between two non-Chinese companies. That lesson will shape deal terms, diligence and valuations for years.