Coverage Guide

Overview

Functions of autonomy like anti-lock braking and cruise control have been developing for decades. Those were safety features to enhance driver ability, not replace the driver. Autonomous Vehicles are wholly new and unique. The insurance marketplace has never seen this risk. Which leads to interesting developments.

First, there is no special auto policy required for an Autonomous Vehicle. Legacy auto insurance policies do not require a human driver to control the vehicle. There’s never been another option! It’s the same reason there weren’t speed limit signs for horses in the 1800s.

Autonomy Vehicle Insurance poses an underwriting hurdle, not a coverage hurdle. See selecting the right carrier for more info on this.

If the auto insurance industry heretofore never contemplated a vehicle without a driver, what else did they miss? How does existing coverage structure apply to a seismic shift in the way vehicles will be driven? We’ll break it down below, covering a brief education and practical application for each line.

Let’s get into the nitty gritty of coverage. How it’s structured and how it needs to improve.

Auto Insurance

What it covers

• Liability because of a Deepath.ai’s use of their vehicle
• Damage to the vehicle caused by an accident

What it doesn’t

• Cyber/hacking related accidents
• Liability for FutureCars’ vehicle, Deepath.ai’s product is a component of their vehicle (OEM/Tier 1 relationship)

Claim example

• Deepath.ai is testing its Chrysler Pacifica on the streets of Mountain View. While in manual mode the safety driver rear ends a vehicle set to make a right turn on a red that stopped short due to cross traffic. Coverage here would apply to the liability associated to damage to the other vehicle, injury to the other driver, and damage to Deepath.ai’s vehicle. (the same scenario would play out if the vehicle was in AV mode. It’s just been the practical experience of the IOA Innovation Group that these accidents happen most frequently when a human is at the wheel).

When to buy

• Day 1. Needed by all Autonomous Vehicle Companies that own vehicles and perform on-road testing or are looking to get a testing permit or register a vehicle.

How it can improve

• Carriers can start to develop auto policy language with autonomous vehicles in mind. There are vehicle differences (equipment stack, electrification) that are currently not accounted for. There are potential challenges to liability and negligence based on ownership and AI drivers. And there are increased telematics and video capabilities to aid in the claims process.

Equipment Stack – Inland Marine

What it covers

• Damage to the equipment stack (LIDAR, RADAR, cameras) caused by an accident or other physical damage.

What it doesn’t

• Liability of any kind.

Claim example

• See the previous claim example (auto). Same circumstances, but the Inland Marine policy responds to the damage to the equipment stack, separate from the damage to the vehicle.

When to buy

• Day 1. All Autonomous Vehicle Companies that own upfitted vehicles need this coverage.

How it can improve

• Carriers insure this exposure in different ways. Some will include it as part of the auto policy. More typical is a separate Inland Marine policy. The downside with the former is policy forms were designed for minimal vehicle upgrades (think running lights and cd decks) and are not wholly applicable to the values of AV Equipment stacks and the cost to upfit/actuate. The latter approach comes with its own deductible, meaning one claim is subject to separate auto and inland marine deductibles. This can be an unnecessary hassle and expense.

Product (General) Liability

What it covers

• Injury to people or property resulting from the failure of Deepath.ai’s product sold or a component of FutureCars’ vehicles.

What it doesn’t

• Automobile liability. See an issue here? An auto accident can be caused by the failure of Deepath.ai’s product. The determining factor is ownership. Deepath.ai’s vehicle (auto claim), or FutureCars’ vehicle (Product Liability)?
• Cyber/hacking related accidents

Claim example

• Deepath.ai enables autonomy for OEM FutureCars, who sells their new model featuring ADAS and autonomy features. The proud new owner of the vehicle engages the feature on the freeway. Everything is going well until the system fails to recognize an emergency vehicle stationary on the shoulder and rams into it, injuring the responders inside. Sound familiar? That’s a product liability claim for Deepath.ai.

When to buy

• As soon as Deepath.ai engages in partnerships where their product is outside of their own control. This is also a contractual requirement for many companies, even if there’s not a strong insurance need.

How it can improve

• The policy is not designed for autonomous vehicles. There are vehicle differences (equipment stack, electrification) that are not accounted for. There are potential challenges to liability and negligence. And there are increased telematics and video capabilities to aid in the claims process.

Technology Errors & Ommissions

What it covers

• Financial loss to a third party (partner, customer, etc.) as the result of the failure of Deepath.ai’s product (software/hardware) to operate as intended

What it doesn’t

• Injury to people or property (see Product Liability).

Claim example

• See the example under Product Liability. That same accident on the roadside that injured people can turn into a large Technology E&O claim. If FutureCars has to recall all models with Deepath.ai’s software they are going to be financially impacted. The Technology E&O policy will respond to their financial loss. For a real world example, look at the airline losses resulting from the Boeing 737 Max recall.

When to buy

• As soon as Deepath.ai engages in partnerships where their product is outside of their own control. Especially critical if Deepath.ai’s partner or customer is revenue dependent on their product/software.

How it can improve

• Did you notice that the same accident can trigger coverage under two separate lines of insurance? What do you think would happen if different insurance companies insured the Product Liability and Technology E&O. Lots of finger-pointing. It’s important to put these coverages with the same carrier, whenever possible. And it will be important to the industry to develop hybrid policies that address the whole risk.

Cyber Liability

What it covers

• Internal costs and external liability following a Cyber incident. This could include loss of Personally Identifiable Information (PII), DoS attacks, Bitlocker shutdowns, Ransomware, etc.

What it doesn’t

• Proactive improvement costs are not covered. Insurance carriers view those costs as mitigation and avoidance measures that are the responsibility of Deepath.ai in order to improve their risk profile.

Claim example

• We kindly refer you to Google News “cyber breach” to see any number of claims examples. They are too frequent to detail here.

When to buy

• As soon as Deepath.ai has employees and partners. This isn’t just coverage for a hack to Deepath.ai vehicles. They are an employer and have PII for employees. When they work with municipalities, partners, research institutes, etc. they have sensitive info in and out of their servers. The work is computer based and Deepath.ai is subject to loss of revenue following an incident.

How it can improve

• Cyber liability is forced to evolve to keep pace with the change in cyber risk. There are constantly new exposures and attack vectors to contend with.

Directors & Officers Liability

What it covers

• Loss in value from mismanagement by executives and the Board of Directors of Deepath.ai. Directors & Officers are not protected in their role managing and positioning the company for success by the corporate veil. D&O insurance covers Deepath.ai and the individuals running it.

What it doesn’t

• There are a variety of exclusions pointing back to other policy types (see above). There are restrictive sublimits of coverage in what has become a hard market, including intellectual property derivative suits, regulatory actions, and biometrics.

Claim examples

• Deepath.ai hires top level engineering talent away from FutureCars. The engineers bring a few pieces of intellectual property with them. FutureCars files suit for theft of Intellectual Property. A correctly structured policy will respond, perhaps at a sublimit.
• Deepath.ai CEO fires off a few late night tweets that affect company value and perception. Stakeholders bring suit for diminution of value.

When to buy

• Ideally Deepath.ai buys this day 1. But the reality is they’ll buy it when investors make it a requirement during late round seed or round A funding. Earlier if Deepath.ai plans on including high profile industry leaders on their Board of Directors, or in advisory roles.

How it can improve

• Price. The market has been hardening for several years. 2020 has seen a sharp rise in cost and reduction in limits and capacity. Tesla recently decided not to renew it’s D&O program. Elon Musk is backing the Directors & Officers personally as an alternative.

Workers’ Compensation

What it covers

• Injury to Deepath.ai workers on the job, without regard to fault or negligence.

What it doesn’t

• Injury during non-work hours (work travel excepted).

Claim example

• An employee of Deepath.ai develops Carpel Tunnel syndrome, prompting physical therapy, surgery, and time off work. Workers’ Compensation takes care of the medical bills, wage replacement, and future medical expenses.

When to buy

• As soon as Deepath.ai hires their first employee. It’s statutorily required in most states.

How it can improve

• Workers’ Compensation is an outdated system that does not always provide efficient access and care. Insurance carriers negotiate with providers to form approved Medical Provider Networks (MPNs). Much like the rest of the U.S. Medical system this creates complications around which providers participate, caseloads, and promptness of resolutions. The legal system built up around Work Comp can also complicate matters (attorneys, litigation, appeals boards, etc.).