Due Diligence in AI: 3 things you need to survive AI scrutiny

AI diligence is a tricky area right now. There are a lot of people that have their antennae up about companies with big claims about how their tools are going to change the world.

And when you're talking about diligence, you're talking about a different playing field than you may be used to because the folks that conduct diligence are almost solely focused on the risks. They don't necessarily balance it out with the positives, the reasons why you're taking these risks.

So from that perspective, you really do have to approach it on their terms. You have to understand, I need to have answers for these folks that speak to the way they assess risk, which is going to be in somewhat of a vacuum. Like, here's the things we care about. Do you have policies, procedures, and evidence that shows that you satisfy our assessment of those risks?

If you want to survive AI due diligence in the M&A market, you need to think about establishing some precedents early on. There are three key areas where you can really make an impact to make this an easy process.

That's adopting some type of tracking program for the tools you're using, adopting policies, and then having some form of oversight to enforce those policies.

• Tracking is pretty straightforward. We just want you to look at your implementation of AI, the different tools you're using, the contracts that govern those tools, and write them down. You want to be able to pull from that in diligence and say, "Here's a clear picture of our usage."
• Then with policies, you should have a reasonable use policy adopted. Somewhere in your employee handbook maybe that is a statement for everybody inside the organization. These are the things we care about, and these are the rules we expect you to follow.
• And then some kind of cross-functional oversight. Maybe this is housed within the C-suite or the legal department, but someone whose responsibility it is to make sure that people are following through with their commitments to the policies.

If you can hit those three notes, you can survive almost any M&A due diligence exercise on your AI implementation.