Artificial Intelligence is optimizing the way businesses function. It is also the next frontier of business liability.

As AI replaces people in manufacturing, medical diagnostics and workplace collaboration, who will be held liable for AI mistakes? Whether it’s an AI systems breach or a device malfunction, the consequences can be serious.

Consider the following scenarios: A hacker infiltrates hospital equipment and changes medication dosages, resulting in serious injury or death. Or, a ‘smart’ building is placed on digital lockdown in an active shooter event. A single AI mishap can mean multiple organizations are physically and financially liable – even when all parties consider themselves “covered.”

It is common for a tech company to procure a technology errors and omissions (E&O) policy to cover financial losses related to a service or product they distribute or utilize. Similarly, businesses seek product liability coverage for software or hardware they create and sell. A cyber policy covers a business for losses they incur at the hand of a data breach. And many businesses procure general liability coverage for traditional slips, trips and falls on their property.

Most of these policies, however, will exclude AI breach or malfunction scenarios like those above, leaving a business with gaping holes in their coverage and no recourse. The following best practices can help minimize liability when engaging AI in your business:

  1. Be proactive about your exposure. Companies that are considering utilizing a particular technology, or are developing a new device, will want to be proactive about their AI exposure. Having a conversation with your broker about the potential risk of the new technology will minimize the chance of finding yourself exposed and without proper coverage. In light of the emerging AI risk, HUB brokers are already engaging in conversations with carriers and underwriters on behalf of specific businesses to ensure coverage gaps are considered before they become an issue, helping businesses contemplate potential exposures before engaging in new AI use.
  2. Scan contracts for indemnification coverage. In many cases, vendor contracts are written in favor of the technology company, and businesses do not fully understand the insurance implications. Have your broker review your vendor contracts for:
    • An indemnification clause. The vendor may already specify in your contract that they will not indemnify you in the incidence of a breach.
    • Limited indemnification. The vendor may specify that indemnification is limited only to the cost of your contract with them. If you have a $4K contract, that’s all you’ll be indemnified for in the event of a malfunction.
    • Flat fee indemnification. Sometimes the vendor contract will specify a flat fee for indemnification.
  3. Look at how the specific technology could potentially affect your business. If an issue is to arise with the technology you’ve implemented, consider what it could be. Will it be a financial loss, affect your supply chain, physical or property damage? Consider the potential resulting business interruption exposure.

Is a Dedicated Artificial Intelligence Liability Policy Next?
Eventually companies will be functioning with a tremendous amount of technology, whether it’s automation in manufacturing, or changes in the way buildings are run or medical care is consumed. Between now and then, underwriters and insurance companies will likely create a dedicated artificial intelligence liability policy, aimed at responding to any and all triggers, regardless of who caused the incident, outage or breach. Until then, contact your HUB broker to discuss how your current coverages can work together to close the gaps on your AI risk.