Lawsuits and Product Liability in AI and Robotics

In the fields of artificial intelligence (AI) and robotics, as with any field, there is consistent pressure to be first to market with unique feature sets or completely new technology. Of course, every brand would love to seize the first-mover advantage. Market ownership brings many rewards, including revenue growth and the goodwill of consumers who will become brand-loyal in the future. At the same time, there is a need to meet safety and compliance objectives that become more complicated over time, especially with technology as complicated as AI and robotics. Compliance and liability concerns may affect your company’s opportunity for a first-mover advantage and force what’s called a delayed-release scenario. If you take too much time to make your AI or robotics product or service compliant, your competitors may enter the market first and grab the lion's share of the revenue, leaving your company fighting for scraps.

On the other hand, if you rush to market, you risk releasing a product that's not fully vetted, potentially jeopardizing the safety of your consumers and the public, and opening up your company to potentially a business-ending liability. Attempting to balance risk with the promise of a first-mover advantage when entering a new market is a difficult task, but one that must be considered. 

Risk Assessment

With the risks mentioned above in mind, it’s important to implement processes and procedures that consider safety and legal risks from the beginning of the development cycle, all the way through to the point the AI or robotics product or service is generally available to customers.

The first step is conducting a risk assessment. Organize regular meetings with key stakeholders to discuss potential threats to the system. It is critical to document what kinds of things could go wrong, how likely are these threats to come to pass, and what kinds of controls could be put into place to mitigate them. Establish and refine a process to ensure your product or service has proper, cost-effective safeguards in place to mitigate the risks identified. Only after this risk management process has been thoroughly executed should you go to market. Knowing that you've not only created the best possible product, but you’ve also implemented completely defensible procedures from a legal perspective, will allow you to defend yourself against potential litigation in the future.

Assembling Your Team

Assembling a cross-functional team of stakeholders to identify threats and mitigation strategies can make or break your privacy and security policy development. The team (sometimes called a strike team or a core team) should include senior members from engineering, of course, because they're developing the product, but you should also have senior members from internal marketing, sales, information security, finance, and legal present at regular meetings to identify threats. Each team brings a unique perspective on priorities and risk tolerance. Without an extensive cross-functional team in place, you could very easily overlook certain threats outside the purview of those planning the AI or robotics product or service.

A Cautionary Tale

Ford Motor Company was sued in the 1970s after its Pinto line of cars sustained accidents involving catastrophic and sometimes fatal rear-end collisions. Issues with the gas tank design made it possible for the fuel tank to ignite and potentially explode. The jury involved in a key case rendered a huge verdict against Ford.

Evidence emerged in the trial showing that Ford was aware of problems with the gas tank. Ford was also aware that it could have installed a part that would have protected the gas tank from potential ruptures in a rear-end accident, but the company made an intentional decision to ignore the threat, thus opening the company up to huge liability risks. Not only that, the part was fairly inexpensive (less than $20, by most reckonings). The cost of that part was very modest in comparison with the benefit that would have brought. Apparently, the jury was very angry at Ford and used the huge verdict to send Ford a message.

It is vitally important that all members of an interdisciplinary team tasked with identifying risks be given an equal voice in determining a course of action for risk mitigation. Companies that gather input from a multitude of voices in their threat-management meetings, especially from its lawyers, can identify and correct risks early, potentially saving the company millions in settlements and reputational damages.

Using examples such as the Ford case, I talk to my clients about the need for a thorough risk assessment and mitigating risk. Clients see the importance of taking time to assess risk and fix problems before a product or service hits the market. Doing the right thing isn’t always the easiest path, but it’s the correct path for customers and for your company.

If threats do eventually materialize, your engineers, technicians, and technical folks are going to be testifying at a trial in front of a jury, accounting for their product development process. An opposing lawyer may cross-examine them about their design decisions in minute detail in front of a jury. While in the witness chair, they will be required to describe in great detail what risks they expected and what they did or didn’t do to mitigate risk.

I tell my clients that they can prepare today to win a lawsuit years from now caused by a future accident involving a product or service they haven’t even finished developing. But they have to start today. And they have to be proactive in preventing risk, and not simply try to fix a product or service on the eve of offering it to customers.