Realtime AI News
Report: Beijing Blocked SoftBank's $5.4B ABB Robotics Deal Over Data Concerns
Beijing reportedly stopped SoftBank's $5.4 billion purchase of ABB's robotics unit out of concern that Chinese robot data could end up training OpenAI's models. The reported rationale reframes a large industrial acquisition as a data-sovereignty question rather than a conventional antitrust review.
Chinese authorities reportedly blocked SoftBank's $5.4 billion acquisition of ABB's robotics business, according to a 24/7 Wall St. report published Wednesday, with the stated concern centering on data rather than market share.
As the report describes it, Beijing's worry was that robotic data generated in Chinese factories and facilities could travel with the acquired business and eventually be used to train OpenAI's models. The deal was stopped on that basis.
That reasoning is notable because it treats industrial sensor and telemetry data as strategic material. Robot fleets generate continuous records of physical work, including motion traces, failure modes and handling patterns, and that data is increasingly the input for embodied AI systems.
It also marks a shift in how outbound investment reviews are being argued. National-security screens have typically focused on technology transfer and critical infrastructure; framing a robotics acquisition as a model-training data question widens the aperture considerably.
The report arrives amid a broader tightening of cross-border dealing in robotics and AI hardware, where data provenance and downstream use have become negotiating issues in their own right rather than compliance footnotes.
Details remain thin. The report does not describe the mechanism of the block, the timing of the review, or whether the parties intend to restructure the transaction. Treat the specifics as reported rather than confirmed.
If the account holds, the practical consequence reaches beyond this one deal. Any acquirer with AI training operations abroad now has to model a new class of objection: not what the technology does, but what the data it touches could eventually teach.
Watch for official statements from Chinese regulators or SoftBank, for any revised structure that separates the data assets from the hardware business, and for whether similar reasoning surfaces in other pending robotics deals.
Why it matters
Moving the review logic from technology transfer to training-data provenance raises the uncertainty around any cross-border robotics asset sale. Buyers may need to address data ownership inside the deal structure up front.
Nearby Updates
All11/06, 02:53
Study finds Anthropic's models improving nearly twice as fast with AI's help
A study reported by Reuters finds that Anthropic's models are improving nearly twice as fast with help from AI. The finding turns a long-running argument about AI accelerating AI research into something closer to a measurable trend, with implications for safety review, competition and regulation.
11/06, 02:02
OpenAI President Greg Brockman says the company could be IPO ready in 2027
OpenAI President Greg Brockman said the company could be ready to go public in 2027, citing accelerating enterprise revenue and a turnaround in its business. TechCrunch's report stops short of a firm listing plan and gives no valuation, underwriters, or exchange.
11/06, 02:00
Claude adds a Personal Finance tool that connects bank accounts through Plaid
Claude now includes a Personal Finance tool that lets users link bank and card accounts through Plaid and then ask questions about their finances. TechCrunch reported the launch on November 5 but gave no pricing, regional availability, or bank coverage details.
11/06, 01:00
OpenAI launches GPT-6 Cyber, a model aimed at cybersecurity work
OpenAI has introduced GPT-6 Cyber, a model built for cybersecurity, with coverage citing better performance alongside stronger safety guardrails. Fortune reported the same launch the same day against growing alarm about AI-driven cyber attacks, signalling that security is now its own product category for frontier model makers.