Hi, everyone, and welcome. And we're live here from New Jersey, Ohio, Asia and Bonna. Are you in California?

Yeah.

California. Alright. So I'm representing the East West Coast and some central there. Alright. So to set the stage, could each of you introduce yourself and share your perspective on the most significant way AI is changing and reshaping the cybersecurity landscape. Let's start with you, Asia.

Sure. So hi. I'm Asia, CIO and CISO here at Treasure Data. I've been with the company for almost six years. But before that, I was with HP, Symantec, Oracle, and some banks, like a point to LLC, IBC. So to your question, how is it changing? Actually, I'm very bullish on AI. And, where people see challenges, I see opportunities. And how is this shaping is, yes, we are talking about how the bad guys are using it for coming up with attack surfaces. I am also seeing it as I said, I'm not only a CSM, also a CIO. The efficiency gains we are having, the productivity hacks we are having, sky is the limit with AI.

So it's a balance that we need to strike and that we will talk about in our chat, innovation and securing at the same time. But, yeah, exciting times.

Fantastic. Over to you, Vonna. Share with us your perspective how AI is reshaping the cybersecurity landscape.

Thanks, Virginia. And, hey. Hi, everybody. I'm Bhavna Singh. I'm the CTO at Okta. Before Okta, I was at, Glassdoor, CTO of Glassdoor. And before that, I worked at few other companies, primarily leading technology. So, when I think about AI, and especially the the Gen AI and the agentic AI that we're talking about these days, I personally am excited about the evolution that we are seeing from AI to gen to to agentic. I think it's I'm also a sci fi fan, so I think it's very, very timely that I've I'm seeing it, and I'm excited that to be living in this era. I will say how it's shaping. I think it's shaping certainly in the way of, you know, opening it up to people who otherwise have been away from tech. So it's really I I I love the fact that it's more inclusive of of as a tech to be more, in the hands of folks who have otherwise not talked about tech or thought about coding and stuff like that.

So I see that as a as a evolution of the whole digitization, but also tech overall. Tech needs to impact everybody, and I see this, further enhancing the value out of of tech. That's how I see it. Of course, there's good part and bad part, and that's why, you know, cybersecurity will talk more about it. But, that's how I see it. Stephanie, over to you.

Awesome. Thank you. So, I'm Stephanie. I, spent the decade of my career doing a lot of ethical hacking, reverse engineering, vulnerability analysis. And so I started out really building that deep appreciation and understanding the hacker mindset. I turned my sights to defense a while ago, and I've done several various things, but I am currently the CSO at Canonical, and we're the creators of the popular Linux operating system, Ubuntu. And though that's the most prominent product we're known for, we've got quite a wide portfolio of really great open source solutions. On the sort of how is AI changing security, so, you know, nobody in the security space, no leader has ever made the comment that there's too much good security talent out there. In fact, we always are at a deficit of security talent.

So I get really excited about AI's ability to fundamentally empower those security amazing people we have to do more. Right? Dig through more data and find more needles and piles of needles. And so I think there's a lot of really interesting opportunity to empower people to do more.

Very good. Okay. So our panel, description mentions AI and how it's redefining the risk risk landscape. So, Bona, as an expert in identity and access management as CTO Okta, can you share with us how AI is impacting identity threats, and what innovative strategies can leaders implement to secure their digital identities in this really evolving landscape?

That's a that's a great question and and a great topic to think about for all of us, not just me because I work at Okta, I would say, primarily because think about AI and agents. What is it? What it is is it's an extension of us doing work on our behalf, and it's multiple of us doing things on our behalf. What that what that means is now there's a technology that's extending your identity, taking your ability to authenticate, authorize, do the work, which also means that I wanna make sure that, I can control it. I can know when to when to when it should stop. I can know that when it's talking to other agents or other systems that it's authenticated, it's right, and and, not a not a bad actor. These are all the things that we are looking into the future as to how will it play the role, and identity is right at the central of it because it's identity that is getting extended.

And how do we make sure that identity is protected, it's governed, it's safe? That will be the meta point that we will discuss when we thought think about AI and agentic AI primarily because that will decide whether we are leaning in, which is are we adopting that technology and how fast we are adopting. Because if we cannot trust it, we can't adopt it. Alright? So that's the big point of how AI and identity is playing a role. That's why the way we see it at Okta and in in many conversation with many leaders across the board, we see identity in AI kind of playing a very, very close contact role if you wanna make sure that the future development around this technology is safe and secure. So that's super, super key. Outside of that, you know, within identity, we also see the fake identity, the fake videos, and all of that. So identity itself of how we authenticate, that also has to raise the bar to make sure that we understand those those capabilities of, you know, biometrics as well.

Like, are we able to make sure that they are real, and and and human identity? So identity itself will have to evolve, identity within agents will have to evolve, and trust will be the key point that will make the decision of adoption, and that's where I see identity playing the key role.

Very good insights. Thank you, Juana. Okay. And a movement that is getting traction in AI and cyber is the convergence of cybersecurity and software architecture, where security consideration is baked into the design and development of systems. Right? Creating more secure and resilient infrastructure. This really shifts us from a post development, you know, security afterthought to a security mindset. I think all of us here is very familiar, of the term secure by design, and it's really, getting stickier in the in the industry. So, Ayesha, in your role as CISO CIO, can you share with us how this convergence is playing out in practice, and what are you seeing as the challenges and the opportunities this presents?

Yeah. So the thing I'll say is this convergence of cybersecurity and software architecture. Now we feel like it's not optional, it's foundational. But ever since I've been in technology for over twenty years, I think that has always been the case. We have been talking about shifting left. What has changed, you're talking about challenges and opportunity, is that before AI, it was IoT or cloud. Now you're only focused on AI per se. It's just getting the limelight. So in our organization, what we have done is we've always talked about building a culture of security innovation at the same time. It needs to run-in parallel. And one thing that I have been, doing ever since I've been with Treasure Data is incentivizing behavior. So for example, for our security engineers, or the developers, what we are doing is we are telling them we will be incentivizing you for you to be coding security.

Whereas for security vehicles security folks, we are telling them, we are going to incentivize you on how do you enable the business. So here, we are making them responsible for innovation. On the other hand, we are making the other party responsible for security. Because before that, we were like security. People are worrying about security and developers are only coding. So that's one shift. But then outside of it, it's the culture only works if top down and bottom up, we are working at the same time, understanding the principles from strategic to tactical level. So those are the major shifts that we are seeing, but I would say the principles haven't changed. It's just that the new attack surfaces are coming, and we just are seeing how do we evolve quickly and put the right guardrails in place.

Mhmm. Anyone want to add to that? Or

I think she's spot on on the and thank you for being the person who says, you know, security people enable, the development, and innovation. That's great collaboration and partnership. I would just wanna kinda plus one to that, and that kind of partnership. I do wanna add to the fact, Aisha, you called out, which is, originally, you you can refer, which is, you know, secure development was a thing from the start. It's just that it's taking more, kinda, you know, spot light primarily because, again, you know, we we all wanna adopt AI fast, but the trust building and, you know, confirmation that it is a data safe and all of that is coming in the way to, like, bring the fear or bring the worry.

So had we bring that secure lifestyle, you know, development life cycle and everything in place, the whole rigor in place, I think that the same basic principle, but it will help us kinda build that, confidence, in our setup.

Absolutely. Yeah.

Very good thoughts. K. Pivoting on over to perhaps one of the hottest, topics in AI is leadership and governance and how AI how they are so crucial and ensure I'm responsible in ethical AI development and deployment. Stephanie, in your role as CISO, can you tell us a bit of, how do you qualify as the key leadership skills and qualities that are most critical for navigating the complexities of AI in cyber?

You're on mute.

You're on mute, Steph.

Sorry. Thank you. So, you know, one of the things, you know, so fundamentally AI is software. Right? So and foremost, it's all of the complexities associated with security of software, and then there's all the extra stuff. Right? So table stakes is all of the complexity of traditional software security. But I think there's a couple extra added elements. So one is a really big emphasis on technical fluency. Right? AI is changing a lot at a very fast pace, and so there's a real need for those who are looking to essentially lead docs on AI and cybersecurity inside of their respective organizations to have a really good ability to keep up with what's changing in the industry. The ability to ask informed decisions about different types of algorithms, different types of usage of AI, that requires a fundamental understanding of how that needle continues to move in the industry.

A risk assessment mindset, which I would say every single person in cybersecurity should have, but again, there's the traditional risk assessment and then there's the new pieces associated with the fact that it is AI. Right? And different questions need to be asked. Right? Different facets come into that risk assessment. Ethical decision making. There is a tremendous amount of complexity be complexity around navigating the privacy implications, the civil liberties, right, around AI system adoption and being able to have those tough conversations and understand and set inside of your organization perhaps the barriers that are appropriate for your company's posture.

There's no right answer here. Right? It's like many things in security, it's about a company deciding what their posture is and what their comfort or acceptance level is. And then just adaptability. Right? The ability to rapidly incorporate new insights from new AI technology developments, incorporate those into your AI strategy, and help build a culture that embraces the adoption of these new technologies, but also is really vigilant in understanding where that line in the sand is on when it's appropriate to use AI and places where it makes sense to do so and empower humans and where there are places that maybe don't make sense to adopt.

That's a good, starting point for us. Aisha, do you want to add on to us? So can you tell us what the essential elements of an effective AI governance could look like, and how can technology leaders ensure their organizations are both innovative and secure. Right? We gotta strike that balance. I I think that's always Stephanie hit on that as well.

Yeah. Yeah. I think I'll go back to that point earlier. If you are starting right now and you're thinking about how do I build a culture of security and then how I build this culture of innovation, you're too late to the king. So and foremost, you have to align on that. Then the piece is is the alignment on top. Many organization that I've been talking to, I see But it's still in discussion phases. Do we need AI? To what level we need AI? Who should be the driver for AI? Should it be the product organization, or is it going to be the CSO organization? Or so a lot of those conversations happening. From my point of view, what we have done internally for treasure data, so we are a CDP of customer data platform.

And like any other product, you have to think about how you're coming up with AI within your product, and then how do you make sure you use AI for the corporate side. So what we have done is is the alignment on top internally with from CEO to all the other c levels. What is going to be AI for us? So drafting our AI ambition. And when I say AI ambition, there are things like, are we only going to use AI for consumers, or are we going to use AI for back office, front office work as well? So in in our case, what we have done is I have taken on the responsibility from a CIO point of view for all back operation works which which we also call enterprise security or enterprise IT. So I am the driver there. And what I've done from a risk governance point of view, four main things.

And fully transparent, we are still figuring things out as we evolve. AI just coming from all different directions, and we are on our toes. So almost a year ago, we only had a policy. Policy to talk about clearly what is allowed, what is not allowed, what tools we can bring into our ecosystem, and what is going to be a clear no. Then the next step was to understanding our own journey. Now that we are seeing we will have product, AI within our product, and AI for our enterprise, are we solving two different problems? So what I did for my organization, I brought in Glean.

At that point, Glean was the answer to solve because what people were doing on the enterprise side looking for information, searches, chats, automation, streamlining. So I was looking at from efficiency gains point of view, whereas my CPO was looking at it from a point of view of how do I make sure that I give more features to our consumers. So that clear distinction really helped. I am going to be responsible here. You're going to be responsible here. And then I had to wear my seesaw hat to make sure as a whole, we are thinking about not AI in different shops, but we have a holistic approach. Now going back to your question that you're asking, what a good framework should entail? Clear policies is a bare minimum. Then on top, data oversight, and that has two pieces. piece is garbage in, garbage out. So how do you make sure that you have the accurate data going in?

Then the piece is, what are we training our models on? Is the data sensitive? What do we do with that? Then the piece to that was risk checks, whether it's models or, systems or integrations. How do we make sure we are regularly checking them for bias, for errors, for risks? And last but not least is the accountability. And accountability going back to as I said, I did not create a vertical, center of excellence. For me, the center of excellence, excellence is horizontal. Because AI is happening in so many different places, I am holding everyone accountable to that framework saying no matter what you do what you are doing from an AI point of view, follow these bare minimum. We have a committee. We have a framework, and we come back and check on each other. How are we doing? Where are we going?

And how are we balancing innovation with risk?

Right. I I totally agree. This has to be an iterative process. It's not a one and done. I think policies, frameworks, governance would continue to evolve depending on the use cases. And Virginie, if I may add As a follow-up, how could what do you oh, yeah. Sure.

I I just wanted to add quick point on the AI governance. I I just wanna call out, you know, I think great points, Aisha made, which is number one, have an AI governance in place. It could look different. It would appear different. It has same high level bullet points of, like, what's okay, what's not okay, and, you know, when it's not okay, what to do and all of that. But but having that in place is number step number one. And number two, I do feel, you know, a lot of people think that governance means more, you know, let's let's say it out loud bureaucracy or, like, red tape and all of that.

But I do wanna say that AI governance and if done right, and if done minimal as well, it just unlocks and and, you know, gives the speed because now you're not going through and worrying about, am I doing it right? Am I using the right tag? Because, you know, some items are approved and some items you need to go check. So it just gives, actually, the team more permission to run faster in some places and puts the right hurdles in places where we should, where, let's say, they're using some critical data of ours, and we wanna make sure that there's hurdle and approval and checks and balances.

So I do wanna say that AI, you know, AI governance, we need to look at it slightly differently. I know there are other aspects of compliance and governance where people think of it more like for the sake of, you know, certification, but this is different.

It is.

Yeah. And how do we from anyone in this panel, how could what advice would you give executive leaders to foster this sort of trust culture to encourage innovation while maintaining accountability and building stronger cybersecurity process? Again, going into that, You know, how do you push the boundaries as engineers and technicians to be innovative while maintaining security?

Yeah. I can go by everyone. If you guys wanna go.

Go ahead, Aisha.

I often say to my organization, trust is the main product, everything else is a wrapper. And in simple English, what I tell my, workforce in Treasure Data is don't you know, like, often what happens is, security organizations are seen as a stick, not as a function of enablement. And I speak to them as as simply as possible. I put myself in a consumer issue. So for example, we all shop in many different places. If I'm a consumer, let's say, of a Target or any brand, and if something bad happens, would I go back and shop there? And now that I have taken that, I create a business value around it by helping them understand why trust matters. Now if we stop going there, they will see shrinking their revenue. Shrinking revenue means less shareholder value.

Less shareholder value means that the business is not gonna make money, which means less bonuses, less incentives, less hiring. So now that I've helped him understand the business value of security, then I come back to building trust. Okay. So now we have understood trust from a security point of view. Trust has another variable there as well, which is how do we make sure that we openly communicate with our employees? We help them understand, what clarity means. We are leading with decisiveness. We are aligned. We are walking. We we walk the talk. It's not only do this, do that. You're responsible for this. There's accountability. There is transparency. So understanding those two angles creates build, builds trust automatically. And then that leads into our processes. I often say successful adoption is not only about technology.

It's about people, processes, and a mindset. And the main focus is the mindset.

For sure.

I think, Aisha, you brought a great perspective from the people, culture, all of that. Let me bring a tech mindset to it. Bring a different angle to the same question, but absolutely aligned with the points that you made. Number one, you know, what's at risk with AI? AI works on data. Do do we have our data properly, you know, tagged or classified to say this is high risk data, too critical for us, or this is low risk. It's okay, you know, to to play with it or or work with it. I think that is important because that's gonna play a role into how fast your team works with it and how confidently they leverage the technology. So that's number one. Number two is, you know, access to the data. So who should have access to the data? What kind of data? How open should it be?

That's where I would say having a foundational identity platform or identity concept and play in place is super key. So if you haven't you know, like Asha said said, you know, if you haven't thought about something, it's too late. I would say if you haven't thought about identity and have it in place and in proper setup, like, yes. You know, that is really something that you should be thinking about because that will lay into who should have access and play into all integrations in future for you. And lastly, I would say, is leaning in into the innovation because now we need to think about our products that we roll out. Not just, you know, we we have had that customer centric or how customer will use it and all of that as a persona. We need AI as a persona.

How will AI interact with our my product now as I'm rolling out? That is a critical concept, a critical shift, or you can say add on that we need as we are thinking about rolling out our products as to not just that humans will use our product now, an AI agent will also use our product. How are we making it easy? Really?

The only thing I'll add then, just, all of those were really great answers. So I'll give one of maybe a unique perspective from us. Right? So everything we do is open source. And one of the things, you know, from my perspective and many of us just culturally, the reason we love open source, right, is open source is capturing the world's innovation as code. And so to us, security and accountability, it's it's part of that mission. It's part of doing good in the world by capturing innovation in code and empowering the world to have this amazing journey together. And for us, it's part of just general quality. Right? Driving that accountability is about building trust and then building building champions.

So we're an organization of about 80 or 90% of our employees are developers. Right? Like, this is we're just fundamentally a technology in the development operation. And so inspiring as a part of that mission, this desire to do good in the world with quality code and a part of that is security. Right? It's a there's a lot of other facets to quality, but security is a really big one. And so that is just a fundamental piece of our culture, right, is making sure that everyone's empowered to understand how to do that correctly. Right? We I'm I'm really fortunate, right, to work in an organization where I spend almost no time convincing people why to do security. Everybody's eager to do it. They want to do the right thing.

So most of my organization's time, right, and my time is spent rather on just giving guidance and empowering tools and methods and things that help those developers be successful because they want to do the right thing. They don't know what the right thing is all the time now. And so fundamentally driving their ability to be successful by giving them those tools, techniques, guidance, has been a really big part of just driving that account. I mean, that accountability is not too hard for us to drive in our culture. People want to do the right thing. It's just empowering them to do the right thing.

Yeah. That must be quite a challenge. But as as you said, Stephanie, it ties back to what Asia said that trust is fundamental in this ecosystem, to to further innovate and and push the boundaries. So really, impactful views, ladies, with regards to the framework and governance around AI. Let's jump over to the ethical considerations for for, AI. We know it's here to stay. We already see its increased use in cybersecurity from threat and anomaly detection, vulnerability management, incident response and recovery, and many more disciplines. It's paramount that organizations ensure responsible and ethical deployment of of AI systems. Stephanie, from your, point of view, how do we balance this need for effective AI driven security measures with concerns about privacy and civil liberties?

Yes. So it's a really good question because there's no easy answer to it. Right? There is no unfortunately, there's no easy answer to this. So there's a couple different aspects I'll bring up. So one is just fundamentally understanding that intersection of AI and civil liberties. Right? Why is this becoming more impactful? So one of the areas is the fact that AI based systems are continuously being used in more, I'll say consequential decision making. Right? They are being used to flag things to say this shouldn't be allowed to happen or this is suspicious. Maybe we should block this thing from happening. So these decisions then fundamentally have an impact on people's freedom of movement, freedom of expression. Right? Just even simple simple like I can or cannot send this message to somebody. Right? So we're seeing an increasingly impactful usage. Right? Decisions being made based on AI.

It's obviously not new news, but I'll still reiterate it that AI trained on historical data often inherits and potentially amplifies bias. Right? Existing societal bias. And so that is again as these AI decisions are you AI engines are used to make these more impactful decisions that potentially have freedom impacting decisions on people. Right? That's something for us to be hyper aware of. Another one that doesn't come up with tremendous amount, but is always interesting to me is just the notion of consent. Right? It is difficult to have meaningful consent for being used in AI agents and AI training and AI usage. And so our notions of your ability to consent for privacy and civil liberties is really really fuzzy with AI systems, right? It's just fundamentally difficult for me to choose whether or not I'm consenting to be trained in some AI.

So these there's this fundamental tension, right, with this desire to use AI in impactful things, but then also talk talk about the training side. Right? So keep in mind that I think it's pretty it goes without saying that an AI is more impactful the more data it is trained on. So fundamentally, there is that tension with civil liberties that to make an impactful AI, I must train it on immense amount of data. Well, I have to get that immense amount of data for somewhere. It It has to involve interesting things. Otherwise, it wouldn't be a super useful AI. So again, there's just a fundamental tension with I can't train AI models arguably without butting up against privacy concerns of where am I getting this immense amount of useful data. So what do we do about it? Right? The things we have to think about are transparency and oversight. Right? I'm a big believer in people making informed decisions.

And so having really clear communication on how data is used, how it's collected, how it's going to, essentially potentially be saved for future usage of AI. Right? I'm also a big believer from a technology perspective that there are technologies emerging that could be really useful in this. So privacy preserving technologies, so things like differential privacy, concepts like federated learning, even honestly use of synthetic data in AI training. I realize it takes an AI to generate synthetic data, but synthetic data being used to then train downstream AI actually is a really powerful use of how we can potentially use AI to actually improve, the potential privacy impact by using things like synthetic data and AI training.

And so all of this to say, right, there's no silver bullet. The key is, you know, where of all of these tensions and making sure that they're in the forefront when the usage of AI, whether you're a company that's developing AI, you're looking to incorporate the usage of AI. Right? These are decisions that will have to be on the forefront of whether or not you use that AI, whether or not you develop that AI, how you distribute that AI to customers, how you use customers data to train your own AIs in house. Right? These all just have to you have to make sure these questions are at the forefront of all of your governance policies. Right? All the great governance things we were just speaking about.

Really, some great, thoughts there, Stephanie. I think we're all in violent agreement that AI will continue to play a strong role in tech, in cybersecurity, cyber resilience. It will help drill drive resilience. And to wrap this up, I mean, maintaining resilience in the future of cyber and given the increasing complexity of cyber threats, organization must build resilience systems and strategies. So and, Vonna, I'll start with you, but I'll open it up to the rest of the panel. Looking ahead, what are some of the most pressing cybersecurity challenges that AI is like likely to exacerbate, and how can leaders prepare for them?

So great question. And I'll tell you the common thing when I when when, you know, people think of this question or topic I hear is, you know, all the deep fakes, all the all the bad actors and all of that. But I will say all of that is true. Yes. But think about that. I think what truly this AI will will exacerbate is the scale of it. The scale of the all of those bad things happening, growing at at a at a, you know, a multi x factor. That is the worry, not just the fact that it's a deep fake or not just the fact that it's a bad attack or not just the fact that it's a it's a bad actor doing something. It's the speed and the scale at which now it can happen. Right? Because it's multiple agents doing multiple things.

They don't have they don't get, you know, tired get tired and all. So that's the number one thing I wanna call out. The number two thing and then I'll I'll, you know, hand it off to my my colleagues here. The number two thing on in terms of what we'll exacerbate is the aspect of of data going out of hands. Because as much as we we we have talked about data governance and all of that, there are many organizations that do not have their data under, you know, right setup and the right guide governance and guidance. So that is also the gonna, will go open up. And that's why I would say, you know, observability and governance is such an important concept. If they were not thought about or for were thought about in a wrong way, I think that it just becomes much and much more, important, and and, rightful in the in the in terms of timing.

I'll hand it off to my panelists here.

Yeah. I think one thought I have is, probably, you have covered all the great points, and I think, those are spot on. Just one thing that I think about, if you think about, organizations who haven't dealt with tech debt, Imagine complexity. She's talking about scalability. Scalability, agility, velocity, complexity, 10 folds. I'm sitting here thinking with my organization, okay, things that we haven't taken care of. How do we do that with agents? And then how do we take care of those agents on top? So so there are a lot of issues, but we don't have all the answers right now. At least there's the awareness that there are issues, and we need to be looking into it. But the points that she made, I think, are spot on.

Yeah. I guess the only thing I'll add is one of the, the pieces that we spend a lot of time internally, thinking about is just the correct spots to use AI. Right? Every time you try to incorporate, right, there's a tremendous amount of government governance oversight, right, decisions that need to be made. And it's you have to make sure that it's use cases that are worth going through all of that for. Right? So one of the things that, like, me in conjunction with our privacy team, our compliance team, right, we spend a lot of time just getting flooded by people who want to try and apply these AI things to just everything. And the truth is we don't have the time to do that. Like, we don't have time to responsibly adopt AI in all of these use cases.

And so a lot more, I think, due diligence needs to go into also considering what are the right use cases where it is worth us going through the act of responsibly adopting. Irresponsible adoption of AI is super easy. Right? It's so easy. The responsible adoption is really hard and we don't have infinite resources to go through and make sure that when we responsibly when we adopt it, it must be responsible to be done. And so trying to find those use cases where it is worth the effort. Right? The value is there to do it the right way and really try to narrow down on those use cases because, yeah, I get hit like left and right of people in different orgs wanting to do this thing with this thing. And it's it's a lot of noise. Right?

We gotta find the right use cases for it where it's impactful and it's worth us putting in the effort to make sure we do it responsibly.

Yeah. One thing that I wanna add to what Stephanie just said, not only identifying the use cases, in my organizations, we've identified 200 use cases which are ready from from an idea point of view. Like, we know that we want those automations to be in place. But the reality and fantasy are two different thing. With AI today, even if I so there is willingness. We want to go all in. But there's technology limitation with AI today where it is. I can only achieve 30 to 40% of those use cases. So that's another reality. It's like not only figuring out what use cases, but also what is feasible.

Really a lot of great points here today, ladies, and a lot to for us to think about. But we're at the end of our panel here, and it really was a pleasure to have this discussion with all of you. Very important, very relevant. Yeah.

Sheila, you're muted. Michelle, you're muted. Hey.