AI Questions to Ask
Bob and I were debating what questions you need to answer when thinking through any new consumer AI product (yes, this is the type of stuff we talk about; I know, I know). To a certain extent, these same questions probably make sense to ask when creating any new startup, product or service but we think these are especially relevant for anything new in AI.
What are the regulatory risks at all levels of United States government — federal, state, municipal, county — and what are the risks and requirements in the EU and other countries?
Which elected officials/agencies have a mandate over the regulation of the product? For consumer products, this means State attorneys general in addition to federal and state subject-matter agencies.
What current laws on the books already regulate the class of product in question, and how do those laws fit with the product vision?
Which existing regulations are most challenging to your product?
Which pending AI regulations at each level of government pose risk to you? (note, while Congress doing much on AI is unlikely, over 800 bills were introduced last year at the state level around AI and that number will increase exponentially in 2025).
Is there an opportunity to use your product’s features to set product standards that will be hard/costly for competitors to meet?
What type of product features are needed to make these regulations feasible?
Is there any opportunity to use legislation or regulation to mandate your product?
Who will oppose your product or the laws/regulations you need?
What groups will lead the fight for rules and laws that are unfavorable to you?
How do you identify them, preempt them, and neutralize them?
What independent groups are potential allies and how can you educate them, align with them, and ultimately mobilize them to help support your regulatory strategy (and tell your story in the media)?
Are there opportunities to mobilize your customers and build grassroots support?
How does your product incorporate reasonable consumer protections?
How do you allay skepticism about big tech and customer privacy?
Where does your product and where does your company fit into the overall AI narrative? In your ideal world, where should AI regulation occur — at the LLM model, at the application level, or are existing laws sufficient?
What are all of the steps necessary–and the costs–of blocking bad regulation or legislation, or of passing good regulations/legislation in each jurisdiction across the United States and globally?
What narrative is necessary to either pass or block relevant regulations?
What’s the timeline, and how far out does the process need to begin (or how long should it be stalled, depending on the strategy)?
Do you want to be out front on the issue, or do you want to create and work through a trade or industry organization?
Are there other questions you would want to answer too? Probably, but if you’re creating a new company – especially an AI company – and you can credibly answer these, you will have done the necessary due diligence to determine whether or not regulatory issues should stop you from moving forward.