To receive future posts by email, please subscribe here.
Show Notes:
Artificial intelligence exclusions are becoming more common, and some of these exclusions are exceedingly broad. This is happening at the same time as the use of artificial intelligence, and thus the exposure associated with artificial intelligence is exploding.
So, risk managers, brokers and other policyholder stakeholders need to be mindful of something we say all the time: Insurance policies are negotiable.
This means the artificial intelligence exclusion your underwriter is trying to slide into your policy may be negotiable.
Let’s talk about a real-world example and why this is necessary.
Consider this exclusion pulled from an actual policy:
"This Policy does not apply to any 'claim,' 'wrongful act,' 'damages' or 'defense costs' based upon, arising out of, or in any way involving any actual or alleged use of 'generative artificial intelligence' by the 'insured.'"
That’s incredibly broad, especially given the “in any way involving” language.
Especially in professional services, media, anywhere where work product is written — the chances that generative AI is used somewhere in the chain is quite high. Would a claim arising out of that work — even if the claim had nothing to do with AI — still “in any way involve” AI? Maybe.
So, what can we do about this?
Well, best case, delete the exclusion.
But, if that’s not possible, reign it in. Replace “based upon, arising out of, or in any way involving” with something like “for.” In other words, is the claim “for” the alleged negligent use of AI? If not, then the exclusion should not apply.
The applicability of an exclusion often comes down to the words, and the message today is: Given the explosion in the use of AI and the risks associated with that explosion in use, be mindful of AI exclusions, try to remove them from your policies, or in the least, try to narrow those exclusions by working with your underwriter to revise the words.