Itochu and Aioi Nissay Dowa to Build US Robotaxi Insurance

Itochu Corporation and Aioi Nissay Dowa Insurance are jointly developing an insurance product built specifically for driverless taxi operators, with sales targeted to begin in the United States as early as fiscal year 2027. The move addresses a gap that has widened as robotaxi fleets scale faster than the insurance products meant to cover them: policies written for human drivers do not map cleanly onto vehicles where the software, not a person, controls braking, lane changes and right-of-way decisions.

Neither company has disclosed pricing, underwriting methodology or the capital structure behind the venture. What is confirmed is the target market, the timeline and the division of labour: Aioi Nissay Dowa brings underwriting and claims expertise from one of Japan's largest non-life insurers, part of the MS&AD Insurance Group, while Itochu brings the trading house's capital, US relationships and experience structuring cross-border mobility ventures. The two companies are betting that a purpose-built product, rather than a retrofitted commercial auto policy, is what robotaxi operators will need as they move from pilot fleets to metropolitan-scale service.

The scale of the two partners underscores how much institutional weight is being put behind the effort. Itochu is one of Japan's five major sogo shosha, trading houses that combine commodity trading, logistics and direct equity investment across sectors from energy to consumer goods to technology, giving it both balance-sheet capacity and a long history of building joint ventures with foreign partners from scratch. Aioi Nissay Dowa sits inside MS&AD Insurance Group, one of Japan's three dominant non-life insurers alongside Tokio Marine and Sompo, with decades of automobile-insurance underwriting experience across both Japanese and international markets. Neither company is a venture-backed insurtech experimenting at the margins; both are large incumbents redirecting core capabilities toward a category that barely existed five years ago.

An Insurance Market Still Built for Human Drivers

Commercial auto insurance in the United States prices risk around driver behaviour: records, hours on the road, claims history. A robotaxi has none of that. Its risk profile depends instead on sensor coverage, software validation, how the vehicle handles edge cases such as unprotected left turns or a cyclist emerging from between parked cars, and how quickly a remote operator can intervene when the system asks for help. Insurers without access to that engineering data are left underwriting blind, which is one reason the market both companies are targeting remains, in their own description, largely undeveloped.

That gap has real consequences for operators. Waymo now runs driverless service across more than a dozen US metro areas, with a fleet that has grown well past a thousand vehicles as it adds cities such as Denver, San Diego and Tampa to an expansion list that already included Phoenix, San Francisco, Los Angeles and Austin. Tesla has pushed its Cybercab and Model Y robotaxi service further into Austin, and Chinese operators including WeRide and Pony.ai are extending right-hand-drive pilots into Singapore and Hong Kong. Every one of those vehicles needs liability coverage, and every operator currently negotiates that coverage without a standardised product built around how autonomous systems actually fail.

The unresolved question sitting underneath all of it is who bears the cost when an autonomous vehicle is at fault: the fleet operator running the service, the software company that built the driving stack, or the manufacturer that built the vehicle. Chinese insurers have already begun answering that question by folding per-ride liability coverage directly into the fare structure for autonomous taxi services, an approach detailed in RobotAIGeek's earlier analysis of how China's insurers priced robot liability. Itochu and Aioi Nissay Dowa are effectively proposing to answer the same question for the US market, where the regulatory and litigation environment is considerably less centralised.

A Product Built on an Existing Telematics Foundation

The robotaxi initiative is not the first time Itochu and Aioi Nissay Dowa have worked together on next-generation vehicle insurance. In 2025, Itochu invested in MOTER Technologies, a US-based insurance-software developer built on telematics and computer-vision data that was established as a wholly owned subsidiary of Aioi Nissay Dowa. MOTER's technology is designed to embed usage-based and risk-scoring insurance software directly into vehicle hardware, an architecture that translates naturally from human-driven electric vehicles to autonomous ones, since both depend on continuous sensor and driving-behaviour data rather than a driver's paper record.

That existing relationship gives the new robotaxi product a technical foundation that a insurer starting from nothing would lack. Aioi Nissay Dowa already has a vehicle-embedded telematics platform through MOTER; Itochu already has the US corporate relationships and capital-markets experience to structure a joint venture. The robotaxi insurance product looks less like a speculative new business line and more like the third step in a deliberate sequence: buy into telematics-based underwriting technology, prove it in the human-driven EV market, then extend the same data architecture to vehicles with no driver at all.

Itochu has also built a broader mobility-data portfolio that supports the same thesis. The trading house has separately partnered with May Mobility to help globalise remote autonomous-vehicle support, giving it operational familiarity with how driverless fleets are actually monitored and managed day to day. Combined with the MOTER stake, Itochu is assembling the pieces of an autonomous-mobility risk business rather than making a single opportunistic bet on one insurance product.

Why Underwriting Discipline Will Decide Who Wins This Market

For robotaxi operators, insurance is not a peripheral cost. At scale, per-vehicle liability coverage becomes one of the largest recurring line items in a fleet's operating budget, alongside vehicle depreciation, remote-operator staffing and compute costs for the driving stack. An insurer that can price risk accurately, using real disengagement and incident data rather than actuarial guesswork borrowed from human-driven fleets, can offer operators materially better terms than a generalist commercial auto carrier applying a blunt surcharge for "autonomous vehicle" as a risk category. That pricing advantage, more than any marketing claim, is what will determine which insurance products robotaxi operators actually choose.

It also means the companies now moving into this space are competing on a narrower and more technical basis than typical insurance competition. Traditional auto insurers compete on distribution, brand and claims service. A robotaxi insurer competes on how well it can model a driving stack's failure modes, which requires either building deep engineering relationships with autonomous-vehicle developers or, as Itochu and Aioi Nissay Dowa have done, acquiring the telematics infrastructure needed to generate that data independently. Operators evaluating coverage will increasingly ask not just about premium levels but about whether an insurer's actuarial models are built on synthetic assumptions or on measured fleet performance.

That distinction should matter to anyone responsible for a fleet procurement or deployment decision, not only to insurers. A robotaxi operator negotiating coverage from a position of data strength, sharing disengagement logs, incident reports and remote-intervention rates directly with an underwriter, is more likely to secure terms that scale sensibly as the fleet grows. An operator negotiating from a position of data weakness, relying on a generalist carrier's borrowed assumptions from human-driven commercial fleets, risks premiums that either overprice genuinely safe operations or underprice genuinely risky ones until a claims event forces a repricing. The nearly 30 robotaxi companies now built on a common perception and compute stack, detailed in RobotAIGeek's mapping of that ecosystem, gives insurers a growing body of comparable technical architectures to underwrite against, which should make the actuarial problem more tractable over time even as the commercial terms remain unsettled today.

The timing also matters competitively. Waymo, Tesla and the Chinese robotaxi operators expanding into new US and Asian markets are all currently self-insuring, using captive arrangements, or working with a patchwork of commercial carriers willing to underwrite autonomous fleets as a favour to a large existing relationship rather than as a standalone product line. A purpose-built insurance product that arrives before regulators or larger US carriers formalise their own frameworks has a real chance of becoming the reference structure that shapes how the entire market prices autonomous-vehicle risk, the same way early standard-setting in cyber insurance shaped how that market matured a decade earlier.

What Determines Whether the Product Actually Ships

Fiscal year 2027 is still more than a year away under the Japanese fiscal calendar that begins in April, and the gap between announcing a joint development effort and bringing a licensed insurance product to market in a foreign jurisdiction is substantial. US insurance regulation is set state by state, meaning Itochu and Aioi Nissay Dowa will need separate approvals in whichever states host their target robotaxi fleets, each with its own review timeline and its own view of how autonomous-vehicle liability should be classified. Neither company has said which states it will target first, whether it will underwrite directly or through a licensed US carrier partner, or how it intends to price coverage relative to the commercial auto policies robotaxi operators use today.

Those details, once disclosed, will show whether this is a genuine attempt to build the reference insurance product for the robotaxi era or a slower-moving joint study that trails the market it is meant to serve. The clearest signal to watch for next is whether Itochu and Aioi Nissay Dowa announce a specific US robotaxi operator as a launch partner before a rival insurer does, since the first standardised policy adopted by a major fleet is likely to become the template every subsequent operator measures its own coverage against.

This analysis synthesizes company disclosures and public reporting on the Itochu and Aioi Nissay Dowa robotaxi insurance initiative as of September 27, 2026. It is for general information purposes only and does not constitute investment, financial, or legal advice.

Hero image credit: Waymo.