Defence Tech Took the Second-Biggest Cheque of the Week: The Insurance Indian Aerospace Startups Cannot Buy Locally

Indian Startup News recorded 13 Indian startup rounds totalling $242.55 million in the week of 10 to 15 August 2026. The second-largest cheque of that week went to Sigma Advanced Systems, a Hyderabad-based company that raised $48.03 million led by Spark Capital, behind only Yulu. The sectors funded that week were semiconductors, healthtech, edtech, EV, foodtech, fintech, homecare, devtools, defencetech and logistics.

That sector list matters more than the individual round. Several of those categories are hardware businesses whose eventual liability attaches to a physical object that somebody else installs, flies, launches or drives. The rest are software or services, where the liability question is professional indemnity and cyber, and where the Indian market has depth. Year to date, Indian startups have raised $8.44 billion across 603 rounds in 2026, and the hardware share of that has grown enough that the insurance gap is a live constraint on delivery contracts rather than a theoretical one.

Founders in this cohort usually discover the gap at the same moment. The first serious customer, a defence prime or a public sector undertaking, sends a draft supply agreement with an indemnity clause and an insurance schedule attached, and the broker comes back saying the Indian market will write a fraction of the limit demanded.

Ordinary product liability cover in India is priced off a fairly stable idea: a defective product injures a person or damages third-party property, and the insurer indemnifies the manufacturer for the resulting legal liability. Indian insurers write that competently for engineering goods, auto components, pharmaceuticals and FMCG.

Aerospace and defence hardware breaks three assumptions inside that structure.

1. The loss is not proportionate to the component. A guidance module worth a few lakh rupees can be the proximate cause of the loss of an airframe worth several hundred crore, plus the crew, plus third-party damage on the ground. Severity is decoupled from turnover, which is the variable Indian product liability rating leans on hardest.
2. The exposure survives delivery by decades. Airframes and platforms stay in service for 30 years or more, and the supplier's liability tail runs with them. A wording written in 2026 has to answer for a component still flying well into the 2050s.
3. Aviation and space risk is a separately underwritten class. Most Indian general insurers' standard liability wordings carry an aviation exclusion, and space exposure is excluded almost universally. Once a component is destined for an aircraft, a satellite, a launch vehicle or a weapons platform, it leaves the general liability market and enters the aviation and space market, which in India is thin and heavily reinsured.

Products liability responds to damage that has already happened. The larger cost in aerospace is usually the damage that has not happened yet.

When a defect is found in one delivered unit, the operator does not wait for a second failure. The fleet is grounded, the affected part is pulled from every airframe or platform carrying it, and the supplier faces claims for inspection, removal, replacement and the operator's loss of use. In aviation programmes this shape of exposure has its own named extensions, commonly written as grounding cover and as products recall or removal cover. Both are optional add-ons, and both are excluded unless bought.

The practical consequences for a hardware startup:

• Grounding liability attaches without a physical loss. No airframe has to be damaged. A regulator's or an operator's decision to withdraw the type from service is enough to crystallise cost.
• The financial exposure scales with the customer's fleet, not the startup's revenue. A supplier with INR 40 crore of annual sales can trigger a grounding across a fleet whose daily revenue exceeds that figure.
• Recall and removal cost is a first-party-shaped expense sitting inside a third-party policy. Insurers sublimit it hard, and the sublimit is often set at a level that looks sensible against turnover and is irrelevant against fleet size.
• Defence platforms have no civil-aviation grounding precedent to price against. Ministry of Defence and prime-contractor agreements handle withdrawal from service contractually rather than through an airworthiness directive, which makes the insurance trigger a drafting question the broker has to resolve before binding.

The discipline is the one that applies across defence manufacturing risk in India: the sublimit is the real limit, and it has to be argued against the customer's exposure rather than the supplier's balance sheet.

Defence and aerospace supply agreements, whether with the Ministry of Defence, a defence public sector undertaking or a private prime, are drafted by buyers with far more bargaining power than a Series A company. Three clause patterns drive the insurance requirement.

Uncapped or high-multiple indemnities. Commercial contracts in most sectors cap the supplier's aggregate liability at contract value or a multiple of fees paid. Defence and aerospace agreements frequently carve death, personal injury, third-party property damage and intellectual property out of any cap, or set the cap above anything the supplier can insure. A founder who signs an uncapped injury indemnity has created an exposure with no ceiling, and then has to buy a ceiling in a market that sells them in fixed increments.

Additional insured and waiver of subrogation. The prime will require naming as additional insured on the supplier's liability policy and a waiver of subrogation in the prime's favour. Both are endorsements the Indian insurer has to agree to explicitly. Neither is automatic, and a policy issued without them fails the contract's insurance schedule even when the limit is correct.

Flowdown of the government's own terms. Where the prime holds a government contract, its terms flow down. The supplier inherits indemnity and insurance obligations drafted for a company several orders of magnitude larger, with no realistic route to negotiate them away once the prime has accepted them upstream.

Send the insurance schedule and the indemnity clause of the draft supply agreement to the broker before commercial terms are agreed. The negotiable version of an uninsurable clause is the draft. After signature, the only remaining option is buying limit at whatever the market charges.

The honest picture for a hardware defence or aerospace startup approaching the domestic market in 2026.

Written locally without difficulty. Fire and property cover on the factory, machinery breakdown, marine cargo on inbound components and outbound deliveries, erection or contract works cover on installation, group health, employee compensation, directors and officers liability at Series A limits, and cyber at moderate limits. None of this is the constraint.

Written locally but thinly. General third-party liability where the aviation exclusion has been deleted or amended. A domestic insurer will often lead a modest limit and then depend entirely on reinsurance support above it. Capacity exists, but the pricing and the wording are set by the reinsurer rather than the Indian lead, and the process runs in weeks rather than days.

Effectively fronted into overseas markets. Aviation products liability at limits that satisfy a prime contractor, grounding cover, aerospace recall and removal, and anything touching launch or in-orbit exposure. The Indian insurer issues the policy for regulatory and contractual purposes, and the risk is ceded almost in full to specialist aviation and space reinsurers. This is a normal structure. It matters to the buyer because the terms come from the reinsurance market, the certification and quality-system questions in the underwriting submission come from that market, and so does the timeline.

Not realistically available. War and terrorism write-back on defence products at meaningful limit, and any cover that would respond to the intended destructive function of a weapons system. The line insurers hold is between the product failing, which is insurable, and the product working as designed, which is not. Founders in the defence segment should assume that line will be drawn and structure the contract around it. The iDEX-funded cohort's version of this problem is the same one seen from the procurement side.

When the required limit exceeds what the onshore market will assemble, GIFT IFSC is the structurally obvious next stop, and it is now large enough to matter.

The Chambers and Partners Insurance and Reinsurance 2026 India guide records around 20 IIOs registered with IFSCA writing reinsurance business from GIFT IFSC, with IIO premium of USD 191.07 million in FY2024-25. An IIO is the IFSC office of a foreign insurer or reinsurer, which means specialty capacity can be reached from an Indian jurisdiction rather than only through a cross-border cession.

What that gives a defence or aerospace startup in practice:

• Specialty underwriters who recognise the risk. Aviation and space products liability is written by a small set of specialist teams globally. Reaching them through an IIO is a shorter path than a domestic insurer's cross-border placement, and the underwriting conversation starts from shared vocabulary.
• USD-denominated limits. Prime contractor and export agreements often specify limits in dollars. A dollar limit placed through GIFT removes the currency mismatch an INR-denominated onshore policy creates when a claim settles abroad.
• A tower the Indian lead can work with. The domestic insurer still issues the policy where the contract or regulation requires Indian paper, and GIFT capacity sits in the reinsurance tower behind it.

The constraints are real. IIOs are reinsurance-focused, so most direct placements still need an onshore fronting insurer and the fronting fee belongs in the budget. Broker capability at GIFT is uneven, and a broker who has never placed through an IIO will quietly route the risk back to the familiar cross-border market. The growth in GIFT IFSC insurance premium is what makes this route worth insisting on when the domestic market has already declined the limit.

Sequenced by when each piece becomes necessary rather than by premium size.

Before the first prototype leaves the building

1. Property and machinery cover on the facility, with the actual replacement cost of test rigs and tooling reflected in the sum insured rather than book value.
2. Marine cargo and transit cover on component imports and prototype movement, including air freight of flight hardware.
3. General third-party liability with the aviation exclusion reviewed line by line, and an explicit written answer on whether the intended end use voids it.

Before the first customer trial or flight test

4. Products liability with an aviation write-back, or a dedicated aviation products policy, at a limit derived from the customer's platform value rather than from turnover.
5. Grounding and products recall or removal extensions, with the sublimit argued against fleet size.
6. Contractual review of every indemnity and insurance schedule in the supply agreement, done by the broker and the lawyer together rather than in sequence.

Before the first commercial delivery contract

7. Additional insured and waiver of subrogation endorsements issued in the prime's exact legal name, with certificates matching the contract's wording word for word.
8. A capacity plan for limit increases, because the second and third customers will each demand more limit than the first, and the market cannot always produce it on the timeline a contract sets.

A defence hardware round the size of Sigma Advanced Systems' $48.03 million buys a production ramp, and a production ramp converts a design risk into a liability tail. Five questions belong on the board agenda in the quarter after a round of that size.

• What is the largest single indemnity the company has signed, and is it capped? If any executed agreement carries an uncapped injury or property indemnity, that is a disclosable exposure regardless of how likely a claim looks.
• Does the current liability policy exclude aviation, and does the product touch an airframe or a launch vehicle? Those two facts together decide whether the company is insured at all for its main line of business.
• What is the grounding sublimit, and what fleet does it face? A sublimit of INR 5 crore against a customer operating 40 platforms is a number chosen for the premium, not for the risk.
• Has GIFT IFSC capacity been tested, or only the domestic market? With around 20 IIOs at GIFT and USD 191.07 million of IIO premium in FY2024-25, a broker who has not looked there has not finished the placement.
• Who owns renewal and limit escalation? Limits adequate at prototype stage are not adequate at rate production, and the increase has to be planned ahead of the contract that requires it.

The same discipline applies across the hardware cohort funded that week. A fabless chip design startup faces a structurally similar products liability problem: a small component, a large downstream consequence, and an exposure sitting in a market the founder has never bought from. The defence and aerospace version is the one where the downstream consequence is largest and the domestic capacity is thinnest.