AI Accountability & Colorado’s Trailblazing AI Consumer Law: Interview with Rep Manny Rutinel [AI Today Podcast] AI transcript and summary - episode of podcast AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion

Go to PodExtra AI's episode page (AI Accountability & Colorado’s Trailblazing AI Consumer Law: Interview with Rep Manny Rutinel [AI Today Podcast]) to play and view complete AI-processed content: summary, mindmap, topics, takeaways, transcript, keywords and highlights.

Go to PodExtra AI's podcast page (AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion) to view the AI-processed content of all episodes of this podcast.

View full AI transcripts and summaries of all podcast episodes on the blog: AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion

Episode: AI Accountability & Colorado’s Trailblazing AI Consumer Law: Interview with Rep Manny Rutinel [AI Today Podcast]

Author: AI & Data Today
Duration: 00:11:50

Episode Shownotes

On May 17, 2024, Governor Jared Polis signed SB24-205, a pioneering law to protect Colorado consumers using AI systems. The legislation mandates transparency and accountability, ensuring AI technologies are developed ethically and fairly. In this episode of AI Today hosts Kathleen Walch and Ron Schmelzer interview Rep Manny Rutinel. He

is a Colorado State Representative and a Prime Sponsor of the bill in the House. Continue reading AI Accountability & Colorado’s Trailblazing AI Consumer Law: Interview with Rep Manny Rutinel [AI Today Podcast] at Cognilytica.

Full Transcript

00:00:01 Speaker_00
The AI Today podcast, produced by Cognolitica, cuts through the hype and noise to identify what is really happening now in the world of artificial intelligence.

00:00:10 Speaker_00
Learn about emerging AI trends, technologies, and use cases from Cognolitica analysts and guest experts.

00:00:22 Speaker_03
Hello, and welcome to the AI Today podcast.

00:00:24 Speaker_01
I'm your host, Kathleen Mulch. And I'm your host, Ron Schmelzer.

00:00:28 Speaker_01
And as AI makes its way into our everyday lives, whether we want it or not, even in some cases, AI technology being embedded in devices that we work with, AI technology that we're interacting with perhaps as customers, or perhaps we're using AI as part of some process in our organization, or maybe we're starting to see it in the public sphere.

00:00:50 Speaker_01
Well, as AI has these impacts, We have to take into account not just sort of the capabilities of what these AI systems can do and their technology and all that sort of stuff.

00:01:00 Speaker_01
We also have to think about all these other aspects on our life because, well, we have machines that are doing things with data that in some cases might be making some very impactful decisions. decisions or using our information different ways.

00:01:12 Speaker_01
And of course, this impacts everything that we've been talking about in our Trustworthy AI series. And if you haven't yet, I encourage you all to listen to our Trustworthy AI series.

00:01:20 Speaker_01
As you know, AI Today has been going on strong since 2017, almost 450 episodes. So we've have had a lot to talk about on this topic. And well, it keeps being a recurring theme for us here on the AI Today podcast.

00:01:33 Speaker_03
Exactly. So we focus a lot on trustworthy AI. I will link to those episodes in the show notes if you haven't listened to those.

00:01:39 Speaker_03
But we also do follow laws and regulations, especially as it relates to data and AI, both at a national level and an international level as well, and state and local.

00:01:50 Speaker_03
Because it really does have far-reaching impacts, artificial intelligence, on everything, our personal and professional lives. So for today's podcast, we're so excited to have with us Colorado State Representative Manny Routinel.

00:02:04 Speaker_03
So welcome, Manny, and thanks so much for joining us.

00:02:07 Speaker_02
Thanks so much, Kathleen and Ron. Excited to be here with you.

00:02:11 Speaker_03
Yeah, Colorado has been doing some wonderful things lately in making the news. We've linked to some of that in our LinkedIn newsletter, and I will put that in the show notes so everybody can listen to it.

00:02:22 Speaker_03
But before we get into that, we'd like to start by having you introduce yourself to our listeners and tell them a little bit about your background and your current role now.

00:02:30 Speaker_02
Sure. Yeah. So as you mentioned at the top, I am a state representative here in Colorado, representing the great people of Adams County. I've been doing this for just over eight months or so, and it's been a real honor.

00:02:44 Speaker_02
It's a tough job, and there's a lot of tough aspects of it, but it's fulfilling in that I get to pass laws like the one we're going to discuss today. So before I was a legislator, I was an environmental attorney.

00:02:56 Speaker_02
My whole job was to sue large corporations that were profiting off of the backs of low income communities of color, like the ones that I've grown up in my whole life.

00:03:07 Speaker_02
And, you know, before that went to law school, grad school, undergrad and all that and worked a bunch of jobs from substitute teacher to McDonald's cashier and as an economist for the U.S. Army Corps of Engineers.

00:03:20 Speaker_02
So I've had a chance to dabble in all sorts of things. And I'll just be remiss if I didn't mention my single mom and the amount of love and care that she put into my life growing up and the miracles that she was able to pull off.

00:03:34 Speaker_02
through some really tough moments. So that's a little bit of my background and excited to get into the details of this new AI bill we just passed in Colorado.

00:03:43 Speaker_01
Yeah, it's fantastic. It's a great background.

00:03:45 Speaker_01
And I really like all of that because it's interesting how, you know, there's lots of technologies that we deal with all the time in our lives, the internet, social, mobile, and of course, all the technologies go even farther back. all of that.

00:03:57 Speaker_01
So let's actually talk about it because one of the interesting things is that Colorado has enacted one of the first in the nation artificial intelligence consumer protection laws.

00:04:06 Speaker_01
So for our listeners that may not be familiar, can you provide a high-level overview of the law and some things to think about?

00:04:13 Speaker_02
Sure. As you mentioned, Ron, AI is slowly taking over more and more aspects of our lives from decisions through applications that we submit for jobs or for universities to the housing, health-related algorithms that decide our fates.

00:04:35 Speaker_02
It's slowly making its way into every aspect of our lives. And in many ways, that's a positive development because these tools are extremely powerful and they can provide greater services than what we've been able to receive in the past.

00:04:49 Speaker_02
And they do it more cheaply and perhaps more effectively. Unfortunately, when it comes to certain decisions, discrimination can rear its ugly head. And that's what this bill is all about.

00:05:04 Speaker_02
It's about making sure that when developers and deployers use these, what we call high-risk artificial intelligence systems, that they do so in a way that doesn't cause algorithmic discrimination, that it does so in a way that doesn't perpetuate the racism, sexism, ableism, and other forms of discrimination from our past.

00:05:25 Speaker_02
And we've got a series of provisions in the bill to help push developers and deployers towards that direction.

00:05:32 Speaker_03
Yeah, for our listeners on the podcast, they know we talk a lot about how data is the heart of AI.

00:05:37 Speaker_03
So you really have to be mindful of the data that you're putting into these systems because you're right, they can perpetuate biases and they can perpetuate areas of society that we don't necessarily want to perpetuate.

00:05:51 Speaker_03
And so we do have to be mindful of that. So it's nice that you're paying attention to this and thinking it's so important that you wanna have this consumer protection law in place.

00:06:01 Speaker_03
What motivated the introduction of this bill and maybe what are some of the specific challenges or concerns that it aims to address specifically in the realm of AI?

00:06:10 Speaker_02
Yeah, Kathleen, as you mentioned, it basically comes down to the data. These algorithms are using the data of our past to be able to make decisions for our future. And there's a truism in this space, and it's garbage in, garbage out.

00:06:26 Speaker_02
Unfortunately, that also can mean racism in, racism out, sexism in, sexism out. And so when we're training these algorithms using the decision-making that we had in our past,

00:06:39 Speaker_02
when there was and there is racism, ableism, sexism, that further perpetuates and perhaps even amplifies those biases into our future.

00:06:50 Speaker_02
And so when it comes to the reasons why we have moved forward with this legislation, and fortunate to say that it has been signed by the governor,

00:07:01 Speaker_02
What we're trying to do is make sure that we allow these artificial intelligence systems to grow, but to do so in a sustainable way. Because if there is racism, sexism, ableism built into these systems, that means that

00:07:17 Speaker_02
there will be a backlash towards using them and folks will revolt against their use entirely. And so we want to make sure that these AI systems are used and continue to make our society more prosperous and profitable.

00:07:29 Speaker_02
But we want to make sure that when we do that, we're not propagating the racism, sexism, ableism from our past.

00:07:37 Speaker_01
Yes.

00:07:38 Speaker_01
So maybe, can you explain some of the details of what's covered in the law and maybe some of the things that the law aims to regulate and kind of maybe who it applies to in terms of the parties that it applies to and sort of the scope of all that?

00:07:53 Speaker_01
Sure.

00:07:54 Speaker_02
So let me give you an example of the sort of AI system that would fall under the scope of this bill. So let's say that you are submitting an application through a web portal.

00:08:07 Speaker_02
And that web portal is tasked with screening your application, your resume, to either throw it in the trash or move it on for further review. Now, what some researchers have found, some researchers have found is that

00:08:26 Speaker_02
Racism can be found even in something as simple as the screening process of a resume. So let's say, for instance, that your name is Jared and you played lacrosse in high school and you list those things on your resume.

00:08:41 Speaker_02
That may mean that you get a boost compared to someone that doesn't have a conventional white name and didn't play a conventional white sport in high school.

00:08:53 Speaker_02
So the sort of sectors that are covered are things like employment decisions, housing decisions. So let's say you're submitting an application for to become a tenant. And perhaps educational enrollment and university opportunities.

00:09:09 Speaker_02
There's a series of other sectors that are covered, but those are some of the bigger ones.

00:09:14 Speaker_02
And the sort of the basic premise of the bill is to make sure that developers and deployers use reasonable care with their high risk artificial intelligence systems to prevent algorithmic discrimination.

00:09:28 Speaker_02
And what the bill does is it lays out what reasonable care can mean.

00:09:34 Speaker_02
by providing what's known as a rebuttable presumption, which is a legal term of art to help provide some semblance of protection for these deployers and developers so that they know what reasonable care should or could meet.

00:09:50 Speaker_02
And so that means things like a risk management policy that the developer implements, that they do impact assessments of their high-risk AI systems, that they do annual reviews of the deployments for these high-risk systems to ensure that these systems aren't causing algorithmic discrimination, that they are notifying consumers.

00:10:11 Speaker_02
when a high risk AI system is making a consequential decision. So those are some of the examples, what we help define as what can provide that rebuttable presumption and that presumption of reasonable care.

00:10:25 Speaker_01
Sounds great. Well, that makes a lot of sense. It's definitely some obligations for developers and for people who are building AI systems to keep that in mind. That's really a good explanation. Thank you, folks.

00:10:36 Speaker_01
We're going to link to the text of the bill so folks can learn more about it and read more about it. Of course, think about how it impacts them. I think that's sort of a good little follow-up here. Can you tell us a little bit about

00:10:48 Speaker_01
sort of some of the aspects of transparency and accountability. You mentioned some of those things. Maybe you could dive into sort of maybe what the expectations are for people who are building these systems or maybe people even who are using them.

00:10:59 Speaker_01
You know, what should we demand or ask from the technology companies who are providing these capabilities, what it does in terms of transparency and accountability?

00:11:11 Speaker_02
Yeah, so on the accountability front, this bill will be enforced by the Attorney General, which in our state is a gentleman by the name of Phil Weiser, who is a true leader in this space and is really interested in making sure that we develop artificial intelligence in ways that are sustainable and don't discriminate and safe and things of that nature.

00:11:33 Speaker_02
And so that's the scope of who is able to enforce the provisions under this law. But this bill does a few different things to make sure that we are transparent in our deployment of artificial intelligence systems.

00:11:47 Speaker_02
One is that deployers who interact with consumers must ensure that these AI systems disclose to the consumers when they're interacting with an AI system. Another thing is that the developers of these AI systems have to provide a statement

00:12:07 Speaker_02
and documentation related to the sort of systems that the developer has created that may engage in what we term high risk behaviors, meaning consequential decisions of the sort that are relevant for algorithmic discrimination.

00:12:23 Speaker_02
There's a lot of sharing that the developer is required to do with the deployer so that the deployer is able to conduct impact assessments and risk management policies and programs and things of that nature.

00:12:36 Speaker_02
And so what this bill is generally about is about making sure that these developers are communicating with the deployers and that the deployers are able to communicate appropriately with the consumers.

00:12:50 Speaker_03
Yeah, so I guess, you know, to understand a little bit more about this bill, is this just for Colorado residents or is this for Colorado companies or anybody that is you know, doing business with and working with people that live in Colorado?

00:13:07 Speaker_03
Maybe can you explain the reach of it?

00:13:10 Speaker_02
Sure. Well, the bill concerns businesses who are doing business in Colorado. But of course, a lot of these businesses aren't just doing business in Colorado, they're doing business across the nation and internationally. And so this is what's

00:13:27 Speaker_02
is often what often occurs in a state like California, where businesses in order to be able to comply with California end up complying with California's provisions for their entire national or international operations.

00:13:38 Speaker_02
And so it's unclear the full scope of

00:13:42 Speaker_02
the bill in terms of all the businesses that it'll cover beyond the ones that are explicitly doing business solely in Colorado, but presumably to be able to comply with these provisions, businesses who work not just in Colorado, but across the nation will hopefully provide this information to all their consumers and all their deployers.

00:14:03 Speaker_03
Yeah, that's great. And I think, you know, a lot of people, a lot of states, a lot of local jurisdictions, even at a federal level, have this wait and see approach about how is the technology going to be used?

00:14:14 Speaker_03
How is it really going to impact people's lives? Because whenever we talk about, you know, laws and regulations, when it comes to artificial intelligence, you can look at it two ways.

00:14:23 Speaker_03
You can either have, you can upfront put laws and regulations in place to say, all right, let's be proactive and say, and do it that way so that you can put those guardrails in place. Or you can say, let's see how the technology is being used.

00:14:36 Speaker_03
And then we're going to put laws and regulations in place to kind of, you know, figure out how best to do that.

00:14:42 Speaker_03
We talk about cell phones, for example, and when you were driving and being able to use your cell phone, it took a long time for those logs and regulations to catch up.

00:14:50 Speaker_03
So I think that we needed to see how it was really being used, what distracted driving really looked like. So the same thing is happening with AI. And so it really is an exciting time.

00:15:01 Speaker_03
As far as this law goes, how do you see this law setting a precedent for other states and potentially even at a federal level to get federal legislation regarding AI laws, specifically around AI governance, AI transparency, and everything there?

00:15:16 Speaker_02
Yeah, so this bill, I think, should serve as a great template for other states to use and expand on.

00:15:24 Speaker_02
So when this bill was originally getting crafted, it was done so alongside some work that was done in Connecticut, alongside a Connecticut senator who was interested in proposing a similar law. Unfortunately, that law didn't go

00:15:40 Speaker_02
all the way, but there's hope that they'll be able to bring that back and that other states across the nation will be able to use our law as a good template to either pass as is or to strengthen, hopefully.

00:15:53 Speaker_02
So you mentioned that, or perhaps you haven't mentioned yet, but the EU AI Act recently has gone into, I'm actually not sure what the exact status of it is, but I know that it's being implemented or in the process of being implemented.

00:16:07 Speaker_02
And I hope that that also serves as a good template for other states across the country. Unfortunately, our federal government is not able to move as quickly on the issues that it needs to at times. And this is probably a great example.

00:16:23 Speaker_02
The artificial intelligence space has been neglected for far too long. And so that's why I think it's important for states to take up that fight to make sure that we protect our residents as much as possible.

00:16:36 Speaker_02
And the beauty of a state like Colorado is that given how large it is, it could make it such that we not only protect Colorado residents, but residents in other states where these companies are also working.

00:16:51 Speaker_02
So yeah, that's the basic hope is that we pass bills like these at the state or local level to hopefully inspire other states and the federal government to replicate or enhance what we've done here in Colorado.

00:17:04 Speaker_03
Yeah, that's great. And it's great you brought up the EU AI Act because we have had some podcasts on that. And so I will link to that in the show notes for our listeners that haven't had a chance to check that out yet.

00:17:14 Speaker_03
But yes, it is, you know, a very comprehensive bill in the European Union where they're saying, I mean, and they do lead the way, you know, with laws and regulations.

00:17:25 Speaker_03
They had GDPR a few years ago, which was really around data and how best to use and manage data. Now they're leading the way with the EU AI Act as well.

00:17:36 Speaker_03
And so, you know, there's different levels of AI, and so they're categorizing it into different levels of risk. So when it comes to the Colorado bill, did you gain any, you know,

00:17:48 Speaker_03
Did you look to the EU AI Act as an inspiration of how you wanted to craft your bill?

00:17:55 Speaker_02
The truth is that I don't think the EU AI Act played that much of a major role in the sort of work that we're doing here.

00:18:04 Speaker_02
But what the Colorado law does is that it allows for developers and deployers to comply with our law so long as they are complying with a law that is similarly as stringent or more stringent than what we've laid out in our provisions.

00:18:23 Speaker_02
And so that means that if a developer or deployer wants to comply with our law, they can comply with the EU AI Act if that act is as strong or stronger than the provisions that we've laid out in our bill.

00:18:38 Speaker_02
But the hope is that we can create uniformity wherever possible. And so that's what that provision allows.

00:18:47 Speaker_01
Fantastic. Well, I think there's a lot of activity going on in this whole space around regulation, around AI, so we're keeping a good eye on it. We're talking to folks like you, who are very much involved in the whole regulatory process.

00:19:01 Speaker_01
There's a lot of conversations happening all around the world, and I think this is kind of the moment. This is kind of where we are at. have seen the impacts of technology in our lives from the internet, as mentioned, mobile and social media.

00:19:13 Speaker_01
And I think this time around, we're trying to get our arms around it before it gets its arms around us. So I think that's a really great thing. And I think that sort of brings up the last question.

00:19:22 Speaker_01
We love to always ask this final question to all the folks who we have on the AI Today podcast. We've been asking this since the very first AI Today podcast. And we love all the perspectives that people provide from their own unique

00:19:34 Speaker_01
way of looking at things. So as a final note, what do you believe the future of AI is in general and its application to organizations, governments and beyond?

00:19:46 Speaker_02
Yeah, that's a great question. And I'm glad to be part of the lineage of folks that get to answer, provide my two cents. I think the best way to describe my position on AI at this stage is a darted optimism.

00:20:01 Speaker_02
I think that AI has the potential to bring about a tremendous amount of prosperity and lift people out of poverty across the world if we deploy it appropriately and sustainably.

00:20:13 Speaker_02
What I'm concerned about are the sorts of issues that I tackled here with Senate Bill 24205 that deals with algorithmic discrimination regarding discrimination based on race, sex, ability, et cetera.

00:20:30 Speaker_02
And I'm also concerned about broader AI safety concerns. As we've mentioned throughout this discussion, these AI systems are becoming increasingly powerful, and they are becoming part of almost all decisions that we're making.

00:20:46 Speaker_02
And they are reaching every corner of the planet, in every server, et cetera.

00:20:53 Speaker_02
And so I'm concerned that either bad actors or actors that are unaware of the consequences of how they use AI systems could end up having really horrible repercussions for our, let's say, our power grids, our bank servers, the ways that we are able to protect ourselves against

00:21:17 Speaker_02
you know, biological weapons or other things of that nature.

00:21:21 Speaker_02
And so I'm concerned about not just the algorithmic discrimination and the ways that that could further entrench inequality in America and across the world, but I'm also concerned about safety considerations related to how these systems can be used to harm us in really catastrophic ways.

00:21:42 Speaker_03
Well, that was a great perspective and way to end. We always get different responses to that question every single time we ask it because you're able to bring in your own unique backgrounds and experiences. So that was a wonderful answer.

00:21:57 Speaker_03
And this has been such a wonderful interview. So thank you for sharing with our listeners what's going on in Colorado. We'll definitely link to it in the show notes.

00:22:05 Speaker_03
And we encourage our listeners to stay engaged and to see what's going on and see how this is going to impact, you know, that broader level. Maybe other states are going to start adopting their own laws.

00:22:15 Speaker_02
Yeah. Kathleen and Ron really appreciate you having me and excited to hopefully come on the next time we pass a bill in Colorado.

00:22:22 Speaker_03
Absolutely. All right, well, thank you so much again. This was such a wonderful podcast. We'll link to everything in the show notes, and I encourage our listeners to check it out.

00:22:31 Speaker_03
Like I said, also subscribe to AI Today if you haven't done so already, so you can get notified of all of our upcoming episodes. And I know that we talked about a lot of different episodes. I'll link to all of those in the show notes.

00:22:42 Speaker_03
I'll also link to our newsletter so that you can check that out as well. Like this episode and want to hear more? With hundreds of episodes and over 3 million downloads, check out more AI Today podcasts at aitoday.live.

00:22:55 Speaker_03
Make sure to subscribe to AI Today if you haven't already on Apple Podcasts, Spotify, Stitcher, Google, Amazon, or your favorite podcast platform. Want to dive deeper and get resources to drive your AI efforts further?

00:23:08 Speaker_03
We've put together a carefully curated collection of resources and tools. handcrafted for you, our listeners, to expand your knowledge, dive deeper into the world of AI, and provide you with the essential resources you need.

00:23:21 Speaker_03
Check it out at aitoday.live slash list. This sound recording and its contents are copyright by Cognolitica. All rights reserved. Music by Matsu Gravas. As always, thanks for listening to AI Today, and we'll catch you at the next podcast.