Matt Mahan

So we publish all that on our website and we're very restrictive.

We don't do facial recognition.

That is an explicit regulation and standard.

We don't keep the data indefinitely.

In fact, with our automated license plate readers, we delete the data.

after 30 days unless it is still actively needed in a prosecution or investigation and it's been kept specifically for that purpose.

Otherwise, it's deleted because we don't want to build long-term profiles on people.

Well, we make them sign a contract with us.

We do not.

So in the case of just the license plate readers, we make sure we own the data so it can't be shared by the company.

They don't control it.

We do.

It's in our server.

Right.

And so I think the importance of being out in front and testing and using these tools in government is so that we don't get blindsided, so that we learn and are in a position to understand the tools so that we can effectively regulate them.

And I think that the state of California over the coming years is going to have to create the regulatory framework around protocols.

privacy, data security, protecting children, all in this new context of AI.

And I think the biggest thing is what you referenced in your New York example.

Transparency is what I think is the most important starting place.

There may be specific things that we need to say, yes, you can do this and no, you can't do that.