WEBVTT - Responsible AI: Why Businesses Need Reliable AI Governance 0:00:02.040 --> 0:00:05.080 Hey, Malcolm Glawell Here, I'm back in your feed today 0:00:05.120 --> 0:00:07.960 because we are re releasing an episode of Smart Talks 0:00:07.960 --> 0:00:13.039 with IBM on a very timely topic, AI governance and 0:00:13.119 --> 0:00:17.880 why regulation is critical to building responsible and accountable AI. 0:00:18.480 --> 0:00:24.320 I hope you enjoy it. Hello, Hello, Welcome to Smart 0:00:24.360 --> 0:00:30.000 Talks with IBM, a podcast from Pushkin Industries, iHeartRadio and IBM. 0:00:30.320 --> 0:00:34.720 I'm Malcolm Glabwell. This season, we're continuing our conversation with 0:00:34.880 --> 0:00:39.360 new creators visionaries who are creatively applying technology in business 0:00:39.360 --> 0:00:43.080 to drive change, but with a focus on the transformative 0:00:43.120 --> 0:00:47.040 power of artificial intelligence and what it means to leverage 0:00:47.080 --> 0:00:51.959 AI as a game changing multiplier for your business. Our 0:00:51.960 --> 0:00:57.160 guest today is Christina Montgomery, IBM's Chief Privacy and Trust Officer. 0:00:57.760 --> 0:01:01.760 She's also chair of IBM's AI FAS Export In addition 0:01:01.840 --> 0:01:06.200 to overseeing IBM's privacy policy, a core part of Christina's 0:01:06.280 --> 0:01:10.640 job involves AI governance, making sure the way AI is 0:01:10.760 --> 0:01:17.039 used complies with the international legal regulations customized for each industry. 0:01:17.720 --> 0:01:22.920 In today's episode, Christina will explain why businesses need foundational 0:01:22.959 --> 0:01:27.520 principles when it comes to using technology, why AI regulation 0:01:27.840 --> 0:01:32.160 should focus on specific use cases over the technology itself, 0:01:32.680 --> 0:01:36.840 and share a little bit about her landmark congressional testimony. 0:01:37.240 --> 0:01:41.440 Last May, Christina spoke with doctor Lori Santos, host of 0:01:41.480 --> 0:01:46.240 the Pushkin podcast The Happiness Lab, a cognitive scientist and 0:01:46.440 --> 0:01:50.720 psychology professor at Yale University. Laurie is an expert on 0:01:50.960 --> 0:01:56.040 human happiness and cognition. Okay, let's get to the interview. 0:01:58.600 --> 0:02:00.680 So, Christina, I'm so excited to talk to you today. 0:02:00.840 --> 0:02:03.040 So let's start by talking a little bit about your 0:02:03.120 --> 0:02:05.840 role at IBM. What does a Chief Privacy and Trust 0:02:05.840 --> 0:02:06.880 Officer actually do. 0:02:07.680 --> 0:02:11.440 It's a really dynamic profession and it's not a new profession, 0:02:12.000 --> 0:02:14.880 but the role has really changed. I mean, my role 0:02:15.000 --> 0:02:18.680 today is broader than just helping to ensure compliance with 0:02:18.840 --> 0:02:23.480 data protection laws globally. I'm also responsible for AI governance. 0:02:23.520 --> 0:02:26.359 I co chair or AI Ethics Board here at IBM, 0:02:26.800 --> 0:02:29.840 and for data clearance and data governance as well for 0:02:29.919 --> 0:02:33.480 the company. So I have both a compliance aspect to 0:02:33.520 --> 0:02:36.200 my role, really important on a global basis, but also 0:02:36.800 --> 0:02:41.679 help the business to competitively differentiate because really trust is 0:02:41.680 --> 0:02:45.400 a strategic advantage for IBM and a competitive differentiator as 0:02:45.400 --> 0:02:49.360 a company that's been responsibly managing the most sensitive data 0:02:49.360 --> 0:02:51.960 for our clients for more than a century now and 0:02:52.120 --> 0:02:54.840 helping to usher new technologies into the world with trust 0:02:54.840 --> 0:02:58.240 and transparency. And so that's also a key aspect of 0:02:58.280 --> 0:02:58.720 my role. 0:02:59.200 --> 0:03:01.480 And so joined us here on smart Talks back in 0:03:01.520 --> 0:03:04.079 twenty twenty one, and you chatted with us about IBM's 0:03:04.080 --> 0:03:07.560 approach of building trust and transparency with AI, and that 0:03:07.680 --> 0:03:09.600 was only two years ago. But it almost feels like 0:03:09.639 --> 0:03:12.480 an eternity has happened in the field of AI since then, 0:03:13.000 --> 0:03:15.360 and so I'm curious how much has changed since you 0:03:15.400 --> 0:03:17.960 were here last time. Were the things you told us 0:03:18.000 --> 0:03:21.080 before you are they still true? How are things changing? 0:03:21.760 --> 0:03:25.239 You're absolutely right, it feels like the world has changed 0:03:25.320 --> 0:03:29.000 really in the last two years. But the same fundamental 0:03:29.040 --> 0:03:33.960 principles and the same overall governance applied to IBM's program 0:03:34.520 --> 0:03:38.880 for data protection and responsible AI that we talked about 0:03:38.920 --> 0:03:41.600 two years ago, and not much has changed there from 0:03:41.640 --> 0:03:44.800 our perspective. And the good thing is we've put these 0:03:44.840 --> 0:03:49.320 practices and this governance approach into place, and we've have 0:03:49.520 --> 0:03:53.200 an established way of looking at these emerging technologies. As 0:03:53.200 --> 0:03:56.640 the technology evolves, the tech is more powerful, for sure. 0:03:56.760 --> 0:04:00.800 Foundation models are vastly larger and more capable and are 0:04:00.880 --> 0:04:04.320 creating in some respects new issues. But that just makes 0:04:04.320 --> 0:04:06.120 it all the more urgent to do what we've been 0:04:06.160 --> 0:04:09.760 doing and to put trust and transparency into place across 0:04:09.800 --> 0:04:12.440 the business to be accountable to those principles. 0:04:13.400 --> 0:04:15.560 And so our conversation today is really centered around this 0:04:15.680 --> 0:04:18.919 need for new AI regulation and part of that regulation 0:04:19.040 --> 0:04:21.920 involves the mitigation of bias. And this is something I 0:04:21.960 --> 0:04:24.320 think about a ton as a psychologist, right, you know, 0:04:24.360 --> 0:04:27.080 I know, like my students and everyone who's interacting with 0:04:27.160 --> 0:04:30.279 AI is assuming that the kind of knowledge that they're 0:04:30.279 --> 0:04:33.320 getting from this kind of learning is accurate, right, But 0:04:33.360 --> 0:04:35.400 of course AI is only as good as the knowledge 0:04:35.400 --> 0:04:37.720 that's going in. And so talk to me a little 0:04:37.720 --> 0:04:40.960 bit about like why bias occurs in AI and the 0:04:41.080 --> 0:04:42.960 level of the problem that we're really dealing with. 0:04:44.040 --> 0:04:47.320 Yeah, Well, obviously AI is based on data, right, It's 0:04:47.440 --> 0:04:52.440 trained with data, and that data could be biased in 0:04:52.440 --> 0:04:54.920 and of itself, and that's where issues could come up. 0:04:54.960 --> 0:04:56.840 They come up in the data, they could also come 0:04:56.920 --> 0:05:00.479 up in the output of the models themselves. So it's 0:05:00.600 --> 0:05:05.320 really important that you build bias consideration and bias testing 0:05:05.440 --> 0:05:08.840 into your product development cycle. And so what we've been 0:05:08.880 --> 0:05:11.440 thinking about here at IBM and doing we had some 0:05:11.520 --> 0:05:14.479 of our research teams delivered some of the very first 0:05:14.480 --> 0:05:17.839 toolkits to help detect bias years ago now right and 0:05:17.880 --> 0:05:21.520 deployed them to open source, and we have put into 0:05:21.560 --> 0:05:24.880 place for our developers here at IBM and Ethics by 0:05:24.920 --> 0:05:28.120 Design playbook that's sort of a step by step approach 0:05:28.880 --> 0:05:35.360 which also addresses very fully bias considerations, and we provide 0:05:35.480 --> 0:05:38.640 not only like here's a point when you should test 0:05:38.680 --> 0:05:41.040 for it and you consider it in the data, you 0:05:41.120 --> 0:05:42.839 have to measure it both at the data and the 0:05:42.880 --> 0:05:46.400 model level or the outcome level, and we provide guidance 0:05:46.440 --> 0:05:49.080 with respect to what tools can best be used to 0:05:49.120 --> 0:05:52.599 accomplish that. So it's a really important issue. It's one 0:05:52.640 --> 0:05:55.600 you can't just talk about. You have to provide essentially 0:05:55.720 --> 0:05:58.960 the technology and the capabilities and the guidance to enable 0:05:59.000 --> 0:05:59.960 people to test for. 0:06:00.600 --> 0:06:03.280 Recently, you had this wonderful opportunity to head to Congress 0:06:03.320 --> 0:06:06.359 to talk about AI, and in your testimony before Congress, 0:06:06.400 --> 0:06:09.400 you mentioned that it's often said that innovation moves too 0:06:09.440 --> 0:06:12.040 fast for government to keep up, and this is something 0:06:12.040 --> 0:06:14.200 that I also worry about as a psychologist, right our 0:06:14.240 --> 0:06:17.160 policy makers really understanding the issues that they're dealing with, 0:06:17.600 --> 0:06:19.880 And so I'm curious how you're approaching this challenge of 0:06:19.920 --> 0:06:22.359 adapting AI policies to keep up with the sort of 0:06:22.440 --> 0:06:24.880 rapid pace of all the advancements we're seeing in the 0:06:24.920 --> 0:06:26.320 AI technology itself. 0:06:27.560 --> 0:06:30.760 I think it's really critically important that you have foundational 0:06:30.800 --> 0:06:36.240 principles that applied to not only how you use technology, 0:06:36.279 --> 0:06:37.800 but whether you're going to use it in the first 0:06:37.800 --> 0:06:39.440 place and where you're going to use and apply it 0:06:39.440 --> 0:06:43.560 across your company. And then your program from a governance perspective, 0:06:43.600 --> 0:06:46.120 has to be agile. It has to be able to 0:06:46.200 --> 0:06:51.920 address emerging capabilities, new training methods, etc. And part of 0:06:51.960 --> 0:06:56.479 that involves helping to educate and instill and empower a 0:06:56.520 --> 0:07:00.320 trustworthy culture at a company so you can spot those 0:07:00.360 --> 0:07:02.640 issues so you can ask the right questions at the 0:07:02.680 --> 0:07:05.880 right time if you try. We talked about during the 0:07:05.920 --> 0:07:10.160 Senate hearing, and IBM's been talking for years about regulating 0:07:10.480 --> 0:07:13.600 the use, not the technology itself, because if you try 0:07:13.600 --> 0:07:17.440 to regulate technology, you're very quickly going to find out 0:07:17.840 --> 0:07:20.760 regulation will absolutely never keep up with that. 0:07:21.000 --> 0:07:23.240 And so in your testimony to Congress, you also talked 0:07:23.240 --> 0:07:26.480 about this idea of a precision regulation approach for AI. 0:07:27.080 --> 0:07:29.280 Tell me more about this. What is a precision regulation 0:07:29.360 --> 0:07:31.720 approach and why could that be so important. 0:07:31.920 --> 0:07:34.880 It's funny because I was able to share with Congress 0:07:35.520 --> 0:07:39.320 our precision regulation point of view in twenty twenty three, 0:07:39.680 --> 0:07:42.480 but that precision regulation point of view was published by 0:07:42.600 --> 0:07:47.280 IBM in twenty twenty. So we have not changed our 0:07:47.400 --> 0:07:52.320 position that you should apply the tightest controls, the strictest 0:07:52.360 --> 0:07:56.600 regulatory requirements to the technology where the end use and 0:07:56.720 --> 0:08:00.720 risk of societal harm is the greatest. So that's essentially 0:08:00.720 --> 0:08:03.680 what it is. There's lots of AI technology that's used 0:08:03.680 --> 0:08:07.320 today that doesn't touch people, that's very low risk in nature. 0:08:07.840 --> 0:08:11.280 And even when you think about AI that delivers a 0:08:11.400 --> 0:08:16.840 movie recommendation versus AI that is used to diagnose cancer, right, 0:08:16.880 --> 0:08:20.360 there's very different implications associated with those two uses of 0:08:20.400 --> 0:08:25.240 the technology. And so essentially what precision regulation is is 0:08:25.240 --> 0:08:28.640 apply different rules to different risks, right, more stringent regulation 0:08:28.920 --> 0:08:32.040 to the use cases with the greatest risk. And then 0:08:32.200 --> 0:08:36.360 also we build that out calling for things like transparency 0:08:36.960 --> 0:08:40.240 you see it today with content right, misinformation and the 0:08:40.440 --> 0:08:44.480 like we believe that consumers should always know when they're 0:08:44.520 --> 0:08:47.679 interacting with an AI system, So be transparent, don't hydro 0:08:47.679 --> 0:08:52.120 your AI. Clearly define the risks. So as a country, 0:08:52.440 --> 0:08:55.319 we need to have some clear guidance right in globally 0:08:55.360 --> 0:08:59.079 as well in terms of which uses of AI or 0:08:59.160 --> 0:09:03.880 higher risk role apply higher and stricter regulation and have 0:09:04.000 --> 0:09:06.600 sort of a common understanding of what those high risk 0:09:06.840 --> 0:09:10.600 uses are and then demonstrate the impact in the cases 0:09:10.640 --> 0:09:15.160 of those higher risk uses. So companies who are using 0:09:15.200 --> 0:09:19.280 AI in spaces where they can impact people's legal rights, 0:09:19.320 --> 0:09:23.679 for example, should have to conduct an impact assessment that 0:09:23.800 --> 0:09:27.560 demonstrates that the technology isn't biased. So we've been pretty 0:09:27.559 --> 0:09:31.680 clear about apply the most stringent regulation to the highest 0:09:31.760 --> 0:09:32.720 risk uses of AI. 0:09:34.160 --> 0:09:36.439 And so so far we've been talking about your congressional 0:09:36.480 --> 0:09:39.320 testimony in terms of, you know, the specific content that 0:09:39.360 --> 0:09:41.840 you talked about, But I'm just curious on a personal level, 0:09:42.160 --> 0:09:44.200 you know, what was that like right like right now 0:09:44.280 --> 0:09:46.520 it feels like at a policy level, like there's a 0:09:46.640 --> 0:09:48.960 kind of fever pitch going on with AI right now. 0:09:49.200 --> 0:09:50.640 You know what did that feel like to kind of 0:09:50.760 --> 0:09:53.160 really have the opportunity to talk to policy makers and 0:09:53.200 --> 0:09:56.200 sort of influence what they're thinking about AI technologies like 0:09:56.280 --> 0:09:57.240 in the coming century. 0:09:57.280 --> 0:10:00.719 Perhaps I was really an honor to able to do 0:10:00.760 --> 0:10:03.640 that and to be one of the first set of 0:10:03.679 --> 0:10:07.839 invitees to the first hearing. And what I learned from 0:10:07.880 --> 0:10:11.200 it essentially is, you know, really two things. The first 0:10:11.200 --> 0:10:15.120 is really the value of authenticity. So both as an 0:10:15.120 --> 0:10:19.160 individual and as a company, I was able to talk 0:10:19.200 --> 0:10:21.679 about what I do. You know, I need a lot 0:10:21.679 --> 0:10:25.520 of advanced prep right. I talked about what my job is, 0:10:26.480 --> 0:10:29.200 what IBM has been putting in place for years now. 0:10:29.800 --> 0:10:33.640 So this isn't about creating something. This was just about 0:10:33.679 --> 0:10:36.000 showing up and being authentic. And we were invited for 0:10:36.040 --> 0:10:38.880 a reason. We were invited because we were one of 0:10:38.880 --> 0:10:43.679 the earliest companies in the AI technology space. We're the 0:10:43.679 --> 0:10:49.120 oldest technology company and we are trusted and that's an honor. 0:10:49.280 --> 0:10:51.520 And then the second thing I came away with was 0:10:51.559 --> 0:10:53.920 really how important this issue is to society. I don't 0:10:53.920 --> 0:10:59.439 think I appreciated it as much until following that experience. 0:11:00.120 --> 0:11:03.840 I had outreached from colleagues I hadn't worked with for years. 0:11:03.920 --> 0:11:06.000 I had an outreach from family members who heard me 0:11:06.040 --> 0:11:09.120 on the radio, you know, my mother and my mother 0:11:09.200 --> 0:11:12.319 in law, and my nieces and nephews and my friends 0:11:12.320 --> 0:11:14.160 of my kids were all like, Oh, I get it. 0:11:14.240 --> 0:11:16.520 I get what you do. Now, Wow, that's pretty cool, 0:11:16.760 --> 0:11:19.560 you know. So that was really probably the best and 0:11:19.640 --> 0:11:21.720 most impactful takeaway that I had. 0:11:22.000 --> 0:11:25.600 The mass adoption of generative AI, happening at breakneck speed, 0:11:26.000 --> 0:11:30.000 has spurred societies and governments around the world to get 0:11:30.080 --> 0:11:36.400 serious about regulating AI. For businesses, compliance is complex enough already, 0:11:36.559 --> 0:11:39.880 but throw anever involving technology like AI into the mix, 0:11:40.280 --> 0:11:46.520 and compliance itself becomes an exercise in adaptability. As regulators 0:11:46.559 --> 0:11:51.120 seek greater accountability in how AI is used, businesses need 0:11:51.200 --> 0:11:56.120 help creating governance processes comprehensive enough to comply with the law, 0:11:56.360 --> 0:11:59.440 but agile enough to keep up with the rapid rate 0:11:59.480 --> 0:12:04.840 of change in AI development. Regulatory scrutiny isn't the only 0:12:04.960 --> 0:12:10.079 consideration either responsible AI governance. A business's ability to prove 0:12:10.160 --> 0:12:14.960 its AI models are transparent and explainable is also key 0:12:15.040 --> 0:12:20.040 to building trust with customers, regardless of industry. In the 0:12:20.080 --> 0:12:24.319 next part of their conversation, Laurie asked Christina what businesses 0:12:24.320 --> 0:12:28.720 should consider when approaching AI governance. Let's listen. 0:12:29.600 --> 0:12:32.000 So it's a particular role that businesses are playing in 0:12:32.080 --> 0:12:34.880 AI governance, Like why is it so critical for businesses 0:12:34.920 --> 0:12:35.840 to be part of this? 0:12:36.720 --> 0:12:41.240 So I think it's really critically important that businesses understand 0:12:41.720 --> 0:12:44.640 the impacts that technology can have, both in making them 0:12:44.640 --> 0:12:48.480 better businesses, but the impacts that those technologies can have 0:12:48.960 --> 0:12:54.079 on the consumers that they are supporting. You know, businesses 0:12:54.200 --> 0:12:58.800 need to be deploying AI technology that is in alignment 0:12:58.960 --> 0:13:00.840 with the goals that they set for it, and that 0:13:00.960 --> 0:13:03.840 can be trusted. I think for us and for our clients, 0:13:04.160 --> 0:13:07.080 a lot of this comes back to trust in tech. 0:13:07.520 --> 0:13:13.200 If you deploy something that doesn't work, that hallucinates, that discriminates, 0:13:13.679 --> 0:13:17.839 that isn't transparent, where decisions can't be explained, then you 0:13:17.960 --> 0:13:21.600 are going to very rapidly erode the trust at best 0:13:21.760 --> 0:13:25.079 right of your clients and at worst for yourself. You're 0:13:25.120 --> 0:13:27.800 going to create legal and regulatory issues for yourself as well. 0:13:27.840 --> 0:13:31.760 So trusted technology is really important, and I think there's 0:13:31.800 --> 0:13:34.000 a lot of pressure on businesses today to move very 0:13:34.080 --> 0:13:36.720 rapidly and adopt technology. But if you do it without 0:13:36.840 --> 0:13:40.319 having a program of governance in place, you're really risking 0:13:40.520 --> 0:13:42.000 eroding that trust. 0:13:41.960 --> 0:13:44.000 And so this is really where I think a strong 0:13:44.040 --> 0:13:47.600 AI governance comes in. Talk about from your perspective, how 0:13:47.840 --> 0:13:51.200 this really contributes to maintaining the trust that customers and 0:13:51.240 --> 0:13:53.120 stakeholders have in these technologies. 0:13:53.320 --> 0:13:55.880 Yeah. Absolutely, I mean you need to have a governance 0:13:55.920 --> 0:13:59.880 program because you need to understand that the technology, particularly 0:13:59.880 --> 0:14:04.520 in the AI space, that you are deploying, is explainable. 0:14:04.600 --> 0:14:08.920 You need to understand why it's making decisions and recommendations 0:14:09.000 --> 0:14:10.440 that it's making, and you need to be able to 0:14:10.440 --> 0:14:12.840 explain that to your consumers. I mean, you can't do 0:14:12.920 --> 0:14:15.000 that if you don't know where your data is coming from, 0:14:15.000 --> 0:14:17.480 what data are you using to train those models, if 0:14:17.480 --> 0:14:21.280 you don't have a program that manages the alignment of 0:14:21.320 --> 0:14:24.880 your AI models over time to make sure as AI 0:14:25.120 --> 0:14:29.520 learns and evolves over uses, which is in large part 0:14:30.120 --> 0:14:33.920 what makes it so beneficial that it stays in alignment 0:14:33.960 --> 0:14:37.440 with the objectives that you set for the technology over time. 0:14:38.080 --> 0:14:41.920 So you can't do that without a robust governance process 0:14:41.960 --> 0:14:45.400 in place. So we work with clients to share our 0:14:45.440 --> 0:14:47.920 own story here at IBM in terms of how we 0:14:47.960 --> 0:14:51.560 put that in place, but also in our consulting practice 0:14:52.320 --> 0:14:56.840 to help clients work with these new generative capabilities and 0:14:56.840 --> 0:14:59.720 foundation models and the like in order to put them 0:14:59.760 --> 0:15:01.840 to work for their business in a way that's going 0:15:01.880 --> 0:15:04.560 to be impactful to that business, but at the same 0:15:04.600 --> 0:15:05.640 time be trusted. 0:15:05.840 --> 0:15:07.640 So now I wanted to turn a little bit towards 0:15:07.680 --> 0:15:11.520 Watson X governance, and so IBM recently announced their AI platform, 0:15:11.600 --> 0:15:15.320 Watson X, which will include a governance component. Could you 0:15:15.320 --> 0:15:17.960 tell us a little more about watsonx dot governance. 0:15:18.560 --> 0:15:20.800 Yeah, I mean before I do that, I'll just back 0:15:20.880 --> 0:15:24.480 up and talk about the full platform and then lean 0:15:24.520 --> 0:15:27.400 into Watson X because I think it's important to understand 0:15:27.680 --> 0:15:33.520 the delivery of a full suite of capabilities, to get data, 0:15:33.720 --> 0:15:36.600 to train models, and then to govern them over their 0:15:36.640 --> 0:15:42.120 life cycle. All of these things are really important. From 0:15:42.200 --> 0:15:45.040 the onset you need to make sure that you have. 0:15:46.000 --> 0:15:50.520 For our watsonex dot AI for example, that's the studio 0:15:50.640 --> 0:15:55.080 to train new foundation models and generative AI and machine 0:15:55.120 --> 0:16:00.880 learning capabilities, and we are populating that studio with some 0:16:01.400 --> 0:16:06.920 IBM trained foundation models, which we're curating and tailoring more 0:16:06.920 --> 0:16:10.200 specifically for enterprises. So that's really important. It comes back 0:16:10.200 --> 0:16:13.360 to the point I made earlier about business trust and 0:16:13.440 --> 0:16:19.840 the need to have enterprise ready technologies in the AI space, 0:16:20.160 --> 0:16:23.680 and then the watsonex dot data is a fit for 0:16:23.800 --> 0:16:27.320 purpose data store or a data lake, and then watsonex 0:16:27.320 --> 0:16:31.960 dot gov. So that's a particular component of the platform 0:16:32.440 --> 0:16:36.160 that my team and the AI Ethics Board has really 0:16:36.240 --> 0:16:39.440 worked closely with the product team on developing, and we're 0:16:39.600 --> 0:16:42.440 using it internally here in the Chief Privacy Office as 0:16:42.480 --> 0:16:46.840 well to help us govern our own uses of AI 0:16:47.000 --> 0:16:52.920 technology and our compliance program here. And it essentially helps 0:16:53.000 --> 0:16:57.640 to notify you if a model becomes biased or gets 0:16:57.640 --> 0:17:00.760 out of alignment as you're using it over time. So 0:17:00.880 --> 0:17:03.480 companies are going to need these capabilities. I mean they 0:17:03.560 --> 0:17:07.760 need them today to deliver technologies with trust. They'll need 0:17:07.800 --> 0:17:11.960 them tomorrow to comply with regulation which is on the horizon. 0:17:11.560 --> 0:17:14.359 And I think compliance becomes even more complex when you 0:17:14.400 --> 0:17:18.560 consider international data protection laws and regulations. Honestly, I don't 0:17:18.560 --> 0:17:21.120 know how anyone on any company's legal team is keeping 0:17:21.200 --> 0:17:23.439 up with us these days. But my question for you 0:17:23.560 --> 0:17:27.160 is really how can businesses develop a strategy to maintain 0:17:27.240 --> 0:17:30.240 compliance and to deal with it in this ever changing landscape. 0:17:30.320 --> 0:17:34.280 It's increasingly more challenging. In fact, I saw statistic just 0:17:34.359 --> 0:17:38.960 this morning that the regulatory obligations on companies have increased 0:17:38.960 --> 0:17:42.840 something like seven hundred times in the last twenty years. 0:17:42.640 --> 0:17:47.240 So it really is a huge focus area for companies. 0:17:47.400 --> 0:17:50.280 You have to have a process in place in order 0:17:50.320 --> 0:17:52.840 to do that, and it's not easy, particularly for a 0:17:52.920 --> 0:17:56.800 company like IBM that it has a presence in over 0:17:56.800 --> 0:18:00.000 one hundred and seventy countries around the world. There's more 0:18:00.000 --> 0:18:04.760 more than one hundred and fifty comprehensive privacy regulations, there 0:18:04.800 --> 0:18:09.320 are regulations of non personal data, there are AI regulations emerging, 0:18:10.320 --> 0:18:14.359 so you really need an operational approach to it in 0:18:14.480 --> 0:18:16.520 order to stay compliant. But one of the things we 0:18:16.560 --> 0:18:18.720 do is we set a baseline, and a lot of 0:18:18.720 --> 0:18:22.159 companies do this as well. So we define a privacy baseline, 0:18:22.200 --> 0:18:27.040 we define an AI baseline, and we ensure then as 0:18:27.040 --> 0:18:29.440 a result of that that there are very few deviances 0:18:29.520 --> 0:18:32.600 because it incorporates in that baseline. So that's one of 0:18:32.640 --> 0:18:34.840 the ways we do it. Other companies, I think are 0:18:34.880 --> 0:18:40.080 similarly situated in terms of doing that. But again, it 0:18:40.240 --> 0:18:42.919 is a real challenge for global companies. It's one of 0:18:42.920 --> 0:18:46.880 the reasons why we advocate for as much alignment as 0:18:46.960 --> 0:18:52.320 possible on the international realm as well as nationally here 0:18:52.359 --> 0:18:56.160 in the US, as much alignment as possible to make 0:18:56.359 --> 0:19:01.320 compliance easier for easier and not just because companies want 0:19:01.359 --> 0:19:04.480 an easy way to comply. But the harder it is, 0:19:04.760 --> 0:19:08.639 the less likely there will be compliance. And it's not 0:19:08.760 --> 0:19:14.840 the objective of anybody, governments, companies, consumers to have to 0:19:15.000 --> 0:19:18.520 set legal obligations that companies simply can't meet. 0:19:18.960 --> 0:19:21.040 So what advice would you give to other companies who 0:19:21.040 --> 0:19:24.000 are looking to rethink or strengthen their approach to AI government. 0:19:24.160 --> 0:19:27.280 You need to start with, as we did, foundational principles, 0:19:27.920 --> 0:19:31.359 and you need to start making decisions about what technology 0:19:31.359 --> 0:19:33.720 you're going to deploy and what technology you're not, what 0:19:33.720 --> 0:19:34.840 are you going to use it for, and what aren't 0:19:34.840 --> 0:19:36.240 you going to use it for? And then when you 0:19:36.280 --> 0:19:40.640 do use it, align to those principles. That's really important. 0:19:40.720 --> 0:19:45.240 Formalize a program, have someone within the organization, whether it's 0:19:45.280 --> 0:19:49.560 the chief Privacy officer, whether it's some other role, a 0:19:49.640 --> 0:19:54.080 chief AI ethics officer, but have an accountable individual and 0:19:54.160 --> 0:19:58.399 accountable organization. Do a maturity assessment, figure out where you 0:19:58.400 --> 0:20:01.159 are and where you need to be, and really start 0:20:01.359 --> 0:20:05.520 you know, putting it into place today. Don't wait for 0:20:05.920 --> 0:20:08.960 regulation to apply directly to your business, because it'll be 0:20:09.040 --> 0:20:09.440 too late. 0:20:10.400 --> 0:20:13.800 So Smart Talks features new creators, these visionaries like yourself 0:20:13.800 --> 0:20:17.360 who are creatively applying technology in business to drive change. 0:20:17.600 --> 0:20:20.280 I'm curious if you see yourself as creative. 0:20:20.720 --> 0:20:24.040 You know, I definitely do. I mean you need to 0:20:24.119 --> 0:20:28.760 be creative when you're working in an industry that evolves 0:20:28.800 --> 0:20:33.320 so very quickly. So you know, I started with IBM 0:20:33.560 --> 0:20:36.680 when we were primarily a hardware company, right and we've 0:20:36.800 --> 0:20:40.199 changed our business so significantly over the years, and the 0:20:40.280 --> 0:20:44.720 issues that are raised with respect to each new technology, 0:20:44.760 --> 0:20:48.400 whether it be cloud, whether it be AI now where 0:20:48.400 --> 0:20:50.000 we're seeing a ton of issues, or you look at 0:20:50.000 --> 0:20:54.159 emergent issues in the space of things like neurotechnologies and 0:20:54.240 --> 0:21:00.840 quantum computers. You have to be strategic and you have 0:21:00.960 --> 0:21:04.200 to be creative and thinking about how you can adapt 0:21:04.840 --> 0:21:10.080 agilely quickly a company to an environment that is changing 0:21:10.119 --> 0:21:11.880 so quickly and. 0:21:11.880 --> 0:21:14.919 With this transformation happening at such a rapid pace. Do 0:21:14.960 --> 0:21:17.040 you think creativity plays a role in how you think 0:21:17.040 --> 0:21:20.359 about and implement, specifically a trustworthy AI strategy. 0:21:22.840 --> 0:21:26.880 Yeah, I absolutely think it does because again, it comes 0:21:26.920 --> 0:21:30.080 back to these capabilities, and there are ways. I guess 0:21:30.280 --> 0:21:34.040 how you define creativity could be different, right, but I'm 0:21:34.080 --> 0:21:37.280 thinking of creativity in the sense of sort of agility 0:21:37.320 --> 0:21:41.400 and strategic vision and creative problem solving. I think that's 0:21:41.680 --> 0:21:44.800 really important in the world that we're in right now, 0:21:44.840 --> 0:21:49.399 being able to creatively problem solve with new issues that 0:21:49.480 --> 0:21:52.560 are rising sort of every day. 0:21:52.960 --> 0:21:54.520 And so, how do you see the role of chief 0:21:54.520 --> 0:21:58.160 privacy officer evolving in the future as AI technology continues 0:21:58.200 --> 0:22:01.080 to advance, Like what stuff should CPOs take to stay 0:22:01.119 --> 0:22:03.040 ahead of all these changes that are come in their way? 0:22:04.560 --> 0:22:08.480 So the role is evolving in most companies, I would 0:22:08.520 --> 0:22:13.480 say pretty rapidly. Many companies are looking to chief privacy 0:22:13.520 --> 0:22:17.040 officers who are ready understand the data that's being used 0:22:17.040 --> 0:22:20.560 in the organization and have programs to ensure compliance with 0:22:20.680 --> 0:22:24.440 laws that require you to manage that data in accordance 0:22:24.480 --> 0:22:27.120 with data protection laws and the like. It's a natural 0:22:27.160 --> 0:22:33.160 place and position for AI responsibility. And so I think 0:22:33.160 --> 0:22:35.840 what's happening to a lot of chief privacy officers is 0:22:35.880 --> 0:22:39.400 they're being asked to take on this AI governance responsibility 0:22:39.400 --> 0:22:42.920 for companies and if not take it on at least 0:22:42.960 --> 0:22:46.399 play a very key role working with other parts of 0:22:46.400 --> 0:22:50.040 the business in AI governance. So that really is changing. 0:22:50.280 --> 0:22:54.280 And if chief privacy officers are in companies who maybe 0:22:54.359 --> 0:22:58.119 haven't started thinking about AI yet, they should, So I 0:22:58.160 --> 0:23:02.000 would encourage them to look at different resources that are 0:23:02.040 --> 0:23:06.399 available already in AI governance space. For example, the International 0:23:06.440 --> 0:23:10.880 Association of Privacy Professionals, which is the seventy five thousand 0:23:10.920 --> 0:23:16.040 member professional body for the profession of Chief Privacy officers, 0:23:16.119 --> 0:23:20.760 just recently launched an AI Governance Initiative and an AI 0:23:20.800 --> 0:23:24.800 Governance certification program. I sit on their advisory board. But 0:23:24.880 --> 0:23:27.560 that's just emblematic of the fact that the field is 0:23:27.640 --> 0:23:29.080 changing so rapidly. 0:23:30.320 --> 0:23:32.880 And so, you know, speaking of rapid change, when you're 0:23:33.200 --> 0:23:35.560 back here on smart Talks in twenty twenty one, you 0:23:35.600 --> 0:23:37.960 said that the future of AI will be more transparent 0:23:37.960 --> 0:23:40.199 and more trustworthy. You know, what do you see the 0:23:40.240 --> 0:23:42.280 next five to ten years holding? You know, when you're 0:23:42.280 --> 0:23:45.200 back on smart Talks in you know, twenty twenty six, 0:23:45.320 --> 0:23:47.000 you know twenty thirty, you know what are we going 0:23:47.040 --> 0:23:49.320 to be talking about when it comes to AI technology 0:23:49.359 --> 0:23:49.960 and governance. 0:23:50.800 --> 0:23:52.959 So I try to be an optimist, right and I 0:23:53.000 --> 0:23:56.840 said that two years ago, and I think we're seeing 0:23:56.880 --> 0:24:01.280