Adding complexity where simplicity never stood a chance. The new proposal for AI liability and Insurance.

In September 2022, the European Commission proposed a directive on adapting non-contractual civil liability rules to artificial intelligence (AILD), with an accompanying impact assessment. The European Parliament's Committee on Legal Affairs (JURI) requested the present complementary impact assessment of the proposal, which focuses on specific research questions. The study, which has just been published as a critique, identifies key areas for improvement in the European Commission's impact assessment, not least an incomplete exploration of regulatory policy options and an abridged cost-benefit analysis, particularly of the strict liability regime. The complementary impact assessment study proposes that the AILD should extend its scope to include general-purpose and other 'high impact AI systems' and software. It also discusses a mixed liability framework that balances fault-based and strict liability.

The insurance sector should keep a very close tab on the AILD as it will reshape the insurance industry in various significant ways, from higher premiums to new product lines specifically tailored for AI, and the sector will have to adapt quickly. Insurers must develop more targeted policies, especially as AI technologies become more integrated into everyday business operations. At the same time, companies that use AI will have to grapple with higher costs and stricter liability rules, making it essential to secure the right coverage to avoid serious financial consequences.

Here’s a simplified overview of how this will potentially affect different aspects of the sector:

1. More Liability, Higher Risks

With this new directive, companies using AI will be held more accountable for the potential harm their systems might cause, like data breaches or discrimination claims. This means that insurers will face more claims involving these risks, especially from industries like healthcare or transportation, where AI systems play a big role. This increase in liability will naturally lead to higher insurance premiums as insurers prepare for a higher likelihood of paying out claims.

2. Need for New Types of Insurance

As AI continues to evolve and become more integrated into everyday business, traditional insurance won’t be enough to cover the new risks AI introduces. Insurers will need to develop specialised insurance policies that specifically cover mistakes in AI systems, privacy breaches, or personal data misuse. These policies must account for the complex nature of AI, which can involve everything from software bugs to more severe issues like biased decision making.

3. Mandatory Insurance for AI

The directive hints that insurance might become mandatory for companies using certain types of AI, particularly in high-risk sectors like autonomous driving or healthcare. Just like auto insurance is compulsory for drivers, companies using AI in critical areas may require insurance to cover any potential harm their systems cause.

4. Bigger and More Complex Claims

One of the most significant changes is the new rules around evidence disclosure, which make it easier for people affected by AI malfunctions to make claims. If AI systems cause harm, it will be simpler for injured parties to sue the companies responsible, potentially leading to an increase in the number of claims and the size of settlements. This will likely make insurance for AI related risks more expensive, as insurers will face a more significant financial burden.

5. Tailored Solutions for AI Systems

As AI systems are diverse and complicated, insurers must create tailored solutions. This means developing new policies for businesses using general purpose AI, such as software that can be adapted for many uses. For instance, an insurance policy covering AI in healthcare would look very different from one covering AI in logistics or finance.

6. Challenges for Smaller Companies

The insurance cost might become a barrier for smaller businesses, especially startups developing AI technology. As premiums increase to reflect the higher risk, small businesses might struggle to afford the necessary coverage, potentially limiting their ability to innovate or scale up. This could create an uneven playing field, where only larger, well established companies can afford to take on AI related risks.

7. Pressure on Reinsurers

Given AI related risks' complexity and potential scale, primary insurers might turn to reinsurers to share the burden. Reinsurers will be crucial in helping spread the risk across multiple players, especially in cases where a single AI malfunction could lead to large scale claims involving various parties.

8. Legal and Market Fragmentation

Without consistent rules across European Union countries, insurance providers might avoid challenges as they adapt to various interpretations of AI liability. This could create confusion and inconsistencies in how policies are priced and what they cover across borders, making it harder for insurance companies to operate smoothly across the EU.

Conclusion

And so, when we thought the insurance world couldn’t get any murkier, with the charming intricacies of cyber security insurance, where terms and conditions seem to be written in disappearing ink, along comes AI liability to sprinkle a little extra confusion. It’s as though the insurance industry took one look at the tangled web of cyber security policies and thought, “You know what this needs? More complexity. And let’s toss in some AI for good measure!”

Introducing AI-specific liabilities will only further elevate the art of fine print, transforming it into something even Houdini couldn’t wriggle out of. If we’ve learned anything from the delightful ordeal of trying to understand whether you’re covered for a cyber breach, it’s that AI will now add another layer of delightful vagueness. Insurers will, no doubt, craft policies that are so artfully ambiguous that you’ll feel you need an AI of your own to interpret them.

So, while businesses scramble to insure themselves against the strange and wonderful risks that AI might bring, it’s comforting to know that the noble tradition of opaque insurance terms remains safe. If anything, it’ll be thriving. After all, why simplify when you can make things even more deliciously complicated?

Article by Dr Ian Gauci

Dr Gauci is the Managing Partner of GTG. This technology-focused corporate and commercial law firm has been at the forefront of significant developments in fintech, cybersecurity, telecommunications, and technology-related legislation.

Disclaimer: This article is not intended to impart legal advice, and readers are asked to seek verification of statements made before acting on them.