Samuel Hill: is your brand new shiny assistant that's supposed to make your life easier, just randomly clicking on stuff that you would not want any of your users to click on. And all of a sudden exposing things that should have never been exposed.
Parrish Gunnels: we're trying to educate on the business side of things like, hey, involve us early. I don't wanna sit on every demo of every product that you're looking at, but If you're getting down to the one or two and you think this is gonna fit your need, then bring us in early.
Samuel Hill: the the day without meetings, that could have been Slack messages. I mean, sign me up. That sounds absolutely amazing. Hello and welcome to Mind What Matters. Another incredible episode for you this week. We're gonna play a clip from our webinar with Parish Gunnels, the CISO of MVB Bank. We talked about AI, we talked about cybersecurity, leadership, all sorts of stuff. And then also we're gonna look at things that are happening in the data security world. A lot of moving and shaking and deals being made, very interesting stuff. And then finally, today, my marketing team is gonna force me to react. To some of their videos for Cybersecurity Awareness Month. Hey, great show for you today. But first, let's get right to the news as we mind the headlines. Okay, it really isn't that much of a surprise that an AI assistant that's tasked with reading through your emails and and helping you in some way, that this AI assistant would be very subject to prompt injection or some kind of malicious input from those emails. So now the phishing attack no longer has to just get the user to click on it. They could also embed things that would make an AI agent or an assistant of some kind take action as well. And now malware Bytes cites in their their research on this, their study on this, that one agent was able to, or something was able to click on a link for things and ended up exposing some some cloud repo keys, AWS credentials, and actual customer database information, all because the the injection made the the the the email made it sound urgent. So it had to take action on it. Now, what can you do about this? A lot of things you can do. Obviously, some training of your agents, some guardrails for that. Maybe don't allow them just to click on things and expose your disk or your stuff, or or maybe having some content in context of what actually is passing through on your email system. but color me not shocked that this is an attack vector now. And but also, you know, when I read it, it did make me pause and think, you know, how am I doing here? What am I accessing? What accesses my emails? Who who has access to that? What agents have access to that? And you know, I did have to go double check a couple things, make sure controls are in the right place. And some stuff was happening in the right way. So one, not surprised by this story at all. but two, also great wake-up call. And and candidly, is this a total from the top-down cybersecurity response that we need? Probably there's some tooling here. however, also I think just as users modeling great behavior, cybersecurity professionals to go and show, hey, no, this is what I'm doing as well to make sure that my emails remain safe from this type of prompt injection. because the last thing you want. is your brand new shiny assistant that's supposed to make your life easier, just randomly clicking on stuff that you would never click on, that you would not want any of your users to click on. And all of a sudden exposing things that should have never been exposed. So personal care, personal caution, make sure we take it. this is just absolutely wild. I immediately think of, the classic movies, The Godfather, or things, making an offer they can't refuse. if you think about it, ransomware gangs, the entire point, they do something terrible to your system and your network, and they're trying to force you into action. And the action they want you to do is to pay them some obscene amount of money as a ransom to get back your data, your systems, and your operation and your business. And now every organization has a line, hopefully. Hopefully your organization has come to a line and an agreement as an executive team, as a board, of where's the break-even point? How much is worth it? You know, do we pay the ransom? Do we not pay the ransom? Are we ready? What do we do here? That's not a decision you can make just in the heat of the battle when the breach notification happens and you're you're responding and you're frantically trying to take care of things. We have to have these decisions made ahead of time. And so if there's a break even point of we know if the ransom is up to this dollar amount. were approved by the board to to pay it or whatever, whatever that strategy is. Very individual, every company, all their, pay this is not a debate about whether you should or should not pay the ransomware demand. That's not the debate here. the debate is what's the line, what's your company's specific policy, and how have you affected that? But the escalation to threats of physical violence to you, to your family, those you care about, beyond the pale. Absolutely, hands down, beyond the pale. Fortunately. dark reading reports that they have not seen this take place or have not heard or or seen reports of this taking place from security leaders. I really hope this has not happened to you, but I would not be surprised if this does indeed happen. Because as companies decide whether they will or will not pay the ransom, sometimes bad things happen and demands are made. And if your company's policy is we're not going to pay, scary world that it might come down to our families. To ourselves, just for the profession we have of trying to keep company data safe from bad guys like this who would be willing to do that. So very sobering story here. if you've heard had this happen to you or have heard of this, by all means let me know. I'd love to chat with you about it because this is unfortunately, I think, a terrifying new reality. One that is unsurprising that we've arrived here, but two, we're gonna have to figure out ways to. be able to make the right decisions for our company in accordance with our guidance and our board and our directions, and also keep us physically and our families, those we care about, safe. it's new reality. Don't love it. Up next, we're gonna play a clip from the webinar we had with Parish Gunnels, the CISO of MVB Bank. Parish is a real AI forward leader, and I truly appreciate that about him. So he he shares some of the things that he's done to build AI into their systems and how they go about it. So here's the clip from last week's webinar that you can go watch of yourself if you care to. I'll leave the link in the show description. Walk me through some of the questions that you're asking as you evaluate. I I loved your one to five rating scale. I mean, I'd love to know
Parrish Gunnels: Yeah.
Samuel Hill: what what a one looks like in your rating scale. And then you know maybe those, you know, threes, fours, potentially even fives.
Parrish Gunnels: Yeah. Well, we we I we're getting to where I it's an ad hoc thing right now. So we're we are trying to move towards a a documented process and flow of how it works, right? But in general, we're trying to educate on the business side of things like, hey, because of the things we've experienced and the things we've seen, involve us early. Like I I don't wanna sit right, I don't wanna sit on every demo that of of every product that you're looking at, but If you're getting down to the one or two and you think this is gonna fit your need, then bring us in early. so we're and then we're asking for a meeting really with their engineers basically as if as much as we can. And say so and we're telling them up front, like we need to understand what is your AI? What are you you know, we're asking, what are you using, right? Are you using Claude? Are you using open AI are you using something else, right? We had an a company the other day said they're using Nova and I and I didn't know what that was. And I went and asked my counterpart and he's like, yeah, that's the stuff AWS has been building and it doesn't have a great reputation. I'm like, okay. So so we're trying to I'd say when we get on the on a call with them, we're we're asking those basic questions about how do you have your LM structured, how is it connected within the platform, etc. We're asking them point blank, did you code? Did you code that service in AI? If you did, can you can you give us a percentage of what you think that is, right? and that usually is followed up with are you do you have life cycle processes in place for development, right? So in other words, have you gone and checked that code? Have you scanned it? Have you you know what have you done?
Samuel Hill: Right. Right.
Parrish Gunnels: then we're asking for some sort of architectural diagram, like to show me how the service the front end platform, if it's client, connected or not, how all that is is geared together. and it's a lot of the QA is a lot of like depending on what how they answer it to, right? and you start getting a feeling of red flags or this seems fairly solid type thing. we also try to have we do have created a center of excellence in inside our organization where these sorts of ideas and things can be submitted because there is a a a an aspect where we can build on our own. and we have a a a great I think we're we're we have been down the journey of you know data centralization and that sort of stuff. So I think that
Samuel Hill: Mm-hmm. Mm-hmm.
Parrish Gunnels: our data lake and those sorts of things and how we handle is is good. And so that make gives me more confidence about building it internally just because it's our licensed AI or LLM that we've we've done. It is our data through our data lake. and the AI team that's doing all that development work is putting an orchestration layer together. There's controls and stuff. so there's just there's just more overall control in order to do that with, right? So there's still there's still a need, I think, for third party services, but I think with AI and some of the risk lending itself to Bringing things back in-house and building it is also a very high consideration, I think, that a lot of companies need to make. we definitely ask them those basic things, and then we also get into hallucinations, model validation, you know, all the quality pieces that, you know, what are you doing to make sure. quality of the response is good. The co what's the confidence level of those responses, right? So we're asking those and and you'll find there's a lot of companies that can't answer those questions, right? So that immediately goes, okay, well, you know, and and we we had one, I'll just share one story. We had one company we were talking with, and we were kind of in our heads putting together the timeline of when they started and where they are today, etc. And based on some other questions they had answered and then we realized that they had probably coded some of the service in OpenAI and then probably at some point use claud code. And then we made the comment to them that, well, okay, if you had coded in Claude code back past, you know, earlier than this date, right, you realize Anthropic came out and said there's problems, right? And you should go back and review and look. And they didn't know that. So that was scary.
Samuel Hill: Okay, a lot of interesting things happening in the cybersecurity market, if you will. There's a lot of the business of cybersecurity. I'm fascinated by it. Obviously, I live and breathe this every single day at work. but interesting in the data security space, we've seen a lot of consolidation. There's a lot of companies that have started and are working their way towards you know, viability, funding, all the things, offering solutions around data security on the wide spectrum of data security. I think we're seeing this across AI security techniques. I think we're seeing this across identity security techniques, things that are really innovating rapidly or being forced to innovate very rapidly in this AI world. And what we're seeing, so one reported and confirmed report is Proofpoint, one of the large DLP companies, very specific, very well known for their email DLP, well trusted company across many organizations, buying Veronis, a one of the early players of. Well-established players in the DSPM space. And this is a this is a very large acquisition, but it's like six billion plus dollars. That's billion with a B. A lot of money. Now, private equity is involved. I think Proof Point is owned by Toma Bravo and then pro and now Veronis, which was in the public market, I believe, is coming into a private equity acquisition. Who knows what's going to happen to these two organizations? the the the thesis, right, for this acquisition play is. That one plus one will equal three or potentially four, if you will, bringing together some DSPM capabilities into the proof point world, into the proof point offering. And also then adding more DLP capabilities, more data security, data in motion capabilities for the Veronist customers in the Veronus world. I don't know what will happen. I know a lot of people are going to be nervous if that's an established, trusted platform in your organization. Acquisitions are hard. It's one of the most difficult things to do in all of businesses. Working your way through an acquisition where what does the roadmap look like? What's the feature delivery look like? Now you have product teams and product organizations that are having to figure out, you know, what resources do we have, what funding do we have, what budget do we have, what's the direction overall? You know, fortunately, these are fairly distinct product lines that should retain a lot of the engineering talent and people to be able to keep moving them forward. So if that's a trusted thing for you, you know, who knows? We obviously pay close attention to it. I know I'll be keeping track of because that's a That's a large amount of money being exchanged to combine a couple of data security, established companies, again in the thesis of one powerhouse company that can really well secure data, both at rest across structured and unstructured data environments, and then in motion, as proof what went off, right? So who knows? hopefully it doesn't go the way of other acquisitions where the complications of all these teens and all this product and this backlog and The the roadmaps and all the stuff, the technical debt that both organizations inevitably hold because every organization has it. How does that all play out? So a lot of questions about that. The other one is CrowdStrike entering the DLP game, if you will, buying an emerging site. Well, sorry, let me back it up a little bit. It's reported that they're in the final stages of this acquisition. Who knows what what will actually end up happening, what the valuation of the acquisition is, what the actual numbers turn out to be. A lot of times though, where there is smoke, there is fire, and you know, CrowdStrike being on the endpoint already, well trusted across so many organizations to provide endpoint security, adding some DLP components, some more advanced DLP DLP components. They've offered that a little bit. They've had some offering in there, but but it's not really taken off. It's not really been a thing that a lot of customers have trusted for a lot of reasons. But being so well entrenched on the endpoint, that's probably the hardest thing to do in DLP is endpoint, the endpoint agent itself. And CrowdStrike obviously has that nailed. So buying a company like Orion for whatever the number ends up being, if it does end up being confirmed and going forward, we'll see. But the business of cybersecurity is just so fascinating because now how do you take this, young, fast-moving startup and incorporate it into the well-entrenched, well-trusted, risk-averse, if you if you believe that, CrowdStrike world. Even though they now have access to all of the endpoints that CrowdStrike has under management, which which totals probably in the billions a lot of endpoint agents out there now that they're gonna try and bring it all together. So, how do you mesh these two things? I don't know. I'm not paid enough to know all that, but I just think it's fascinating to see the consolidation because data security has become a hyper focused area, especially as AI is so rampantly running through organizations, it's touching all the data. Now it becomes a DLP problem, a data security problem of how do we keep data safe in this new AI world. So keep your eyes open. If you're customers in any of these places, obviously questions to ask of your vendors and of your partners, figure out, hey, what's the blast radius from this acquisition in my organization and what am I doing here? And then what's our strategy going into next year to either mitigate that blast radius, take advantage perhaps of some economies of scale or whatever those things are, but just new realities in the world of data security. I expect we'll see more. I do expect we'll see more consolidation as companies kind of try to figure out trying to get their hands on how do we secure the data that AI uses. Hey, it is Cybersecurity Awareness Month. And so all over security teams are finding ways to make their users aware of cybersecurity concerns. You know, nothing beats the front line where a user can spot the weird sounding written email or the, you know, know that this is not the right thing to do, being aware of best cybersecurity practices. And also, it's a great time of year to remember and celebrate and honor those women and men who are working so hard to keep data safe. And also for cybersecurity companies, their marketing teams also work extra hard to try and deliver things that are are engaging and fun for it. So, in that spirit, my marketing team is forcing me, hands tied behind my back, to watch videos that they've put together for Cybersecurity Awareness Month. So We're gonna watch them together. I'm gonna react to them and well, let's see what they've got in store for me. yeah, all the all the time. Okay. All right. Yeah, hey, I like that one. yeah, taking the time of user education. It's something that, everybody does all year and in in this month particularly. And how much you have to educate, especially in this AI world where AI is such a prevalent thing and users are doing all sorts of stuff. but I love the juxtaposition now. It's you get to do Oktoberfest instead of regular other stuff for October, the month of cybersecurity awareness month. So I thought that was kinda I thought that was kinda cute. All right, let's take a look at this second video. Yeah, all all the time. That happens. man, hey, I I like that one too. I mean, obviously something you educate users on is you don't share that file, that sensitive thing to everyone with a link, and yet they still do, and we haven't educate all the time. I'm not sure about the Yankees, man. Like I would imagine you're either you either love the Yankees or you really don't love the Yankees. And a lot of people fall into that. I don't love the Yankees camp. So but I guess there's a little, you know, silver lining. The Yankees got bounced from the playoffs. so they are no longer playing October baseball. So I I don't know. But either way, a a cute video. again, the reward of being able to watch your team in October versus having to do so much user education on stuff. Yeah, good job. All right, let's watch this third video together. Wait, there's more. Finally, one entire day without meetings that could have been slacking. that'd be a gift. With mine, you can automate many of the tedious data security. Projects include rested analysts, fewer single points of failure, and a CISO with newfound empathy. Donuts sold separately. Let's mind what matters. Alright, that's that's kinda that's kinda cute. I I mean I guess what would a team really want, you know, instead of a pizza party, you know, which a lot I no, nothing wrong with pizza, honoring your team with some some great lunch or food or things like that. But but yeah, hey, the the day without meetings, that could have been Slack messages. I mean, sign me up. That sounds absolutely amazing. it sounds really great, great video. That was I really like that one. All right, let's watch this last video together. And so the board presentation is not tagged as confidential and slips right past the tools meant to protect it. Looks like another successful user training. Mind can automatically apply sensitivity labels to your data so you no longer rely on users. Or do it for them. This October, you can admire the foliage instead of retraining people. Be aware this October, not anxious. man, that's the classic, right? The annual user training where you you they have the videos of the online content they have to go through and they they they pass the knowledge checks and a test for it and it's great, man. We gotta do it. gotta hit those marks for all the compliance reasons and everything. everybody everybody does it. The completion rate's really high. But were they actually watching it? Did they actually make a difference? And so I love gosh, I love Bob. We all know Bob, right? He's got definitely somebody we all know. And I love the ending where they know now I get to go look at foliage and be see the be outside and enjoy the fall time and and all that kind of stuff. That was pretty. I would love to go hiking there. It looks looks great. So great job. congrats, mind marketing. Those are great videos. I think they really summate the problem really well of of how hard it is to train users to keep them on the straight and narrow, if you will, in cybersecurity best practice and and some fun, cool videos as well that kind of help tell that story. Well, hey, all right, that's gonna do it for another episode of Mind, What Matters. I really appreciate you hanging out with us this week. And as always, if you could like and subscribe to the podcast so you can get notified of new episodes and really just helps us be found by other cybersecurity practitioners who want to maybe put a little more humanity in their day-to-day. Hey, with all that to be said, I'm really grateful for your time. My name is Samuel Hill, and that's all for now.
Parrish Gunnels: we're trying to educate on the business side of things like, hey, involve us early. I don't wanna sit on every demo of every product that you're looking at, but If you're getting down to the one or two and you think this is gonna fit your need, then bring us in early.
Samuel Hill: the the day without meetings, that could have been Slack messages. I mean, sign me up. That sounds absolutely amazing. Hello and welcome to Mind What Matters. Another incredible episode for you this week. We're gonna play a clip from our webinar with Parish Gunnels, the CISO of MVB Bank. We talked about AI, we talked about cybersecurity, leadership, all sorts of stuff. And then also we're gonna look at things that are happening in the data security world. A lot of moving and shaking and deals being made, very interesting stuff. And then finally, today, my marketing team is gonna force me to react. To some of their videos for Cybersecurity Awareness Month. Hey, great show for you today. But first, let's get right to the news as we mind the headlines. Okay, it really isn't that much of a surprise that an AI assistant that's tasked with reading through your emails and and helping you in some way, that this AI assistant would be very subject to prompt injection or some kind of malicious input from those emails. So now the phishing attack no longer has to just get the user to click on it. They could also embed things that would make an AI agent or an assistant of some kind take action as well. And now malware Bytes cites in their their research on this, their study on this, that one agent was able to, or something was able to click on a link for things and ended up exposing some some cloud repo keys, AWS credentials, and actual customer database information, all because the the injection made the the the the email made it sound urgent. So it had to take action on it. Now, what can you do about this? A lot of things you can do. Obviously, some training of your agents, some guardrails for that. Maybe don't allow them just to click on things and expose your disk or your stuff, or or maybe having some content in context of what actually is passing through on your email system. but color me not shocked that this is an attack vector now. And but also, you know, when I read it, it did make me pause and think, you know, how am I doing here? What am I accessing? What accesses my emails? Who who has access to that? What agents have access to that? And you know, I did have to go double check a couple things, make sure controls are in the right place. And some stuff was happening in the right way. So one, not surprised by this story at all. but two, also great wake-up call. And and candidly, is this a total from the top-down cybersecurity response that we need? Probably there's some tooling here. however, also I think just as users modeling great behavior, cybersecurity professionals to go and show, hey, no, this is what I'm doing as well to make sure that my emails remain safe from this type of prompt injection. because the last thing you want. is your brand new shiny assistant that's supposed to make your life easier, just randomly clicking on stuff that you would never click on, that you would not want any of your users to click on. And all of a sudden exposing things that should have never been exposed. So personal care, personal caution, make sure we take it. this is just absolutely wild. I immediately think of, the classic movies, The Godfather, or things, making an offer they can't refuse. if you think about it, ransomware gangs, the entire point, they do something terrible to your system and your network, and they're trying to force you into action. And the action they want you to do is to pay them some obscene amount of money as a ransom to get back your data, your systems, and your operation and your business. And now every organization has a line, hopefully. Hopefully your organization has come to a line and an agreement as an executive team, as a board, of where's the break-even point? How much is worth it? You know, do we pay the ransom? Do we not pay the ransom? Are we ready? What do we do here? That's not a decision you can make just in the heat of the battle when the breach notification happens and you're you're responding and you're frantically trying to take care of things. We have to have these decisions made ahead of time. And so if there's a break even point of we know if the ransom is up to this dollar amount. were approved by the board to to pay it or whatever, whatever that strategy is. Very individual, every company, all their, pay this is not a debate about whether you should or should not pay the ransomware demand. That's not the debate here. the debate is what's the line, what's your company's specific policy, and how have you affected that? But the escalation to threats of physical violence to you, to your family, those you care about, beyond the pale. Absolutely, hands down, beyond the pale. Fortunately. dark reading reports that they have not seen this take place or have not heard or or seen reports of this taking place from security leaders. I really hope this has not happened to you, but I would not be surprised if this does indeed happen. Because as companies decide whether they will or will not pay the ransom, sometimes bad things happen and demands are made. And if your company's policy is we're not going to pay, scary world that it might come down to our families. To ourselves, just for the profession we have of trying to keep company data safe from bad guys like this who would be willing to do that. So very sobering story here. if you've heard had this happen to you or have heard of this, by all means let me know. I'd love to chat with you about it because this is unfortunately, I think, a terrifying new reality. One that is unsurprising that we've arrived here, but two, we're gonna have to figure out ways to. be able to make the right decisions for our company in accordance with our guidance and our board and our directions, and also keep us physically and our families, those we care about, safe. it's new reality. Don't love it. Up next, we're gonna play a clip from the webinar we had with Parish Gunnels, the CISO of MVB Bank. Parish is a real AI forward leader, and I truly appreciate that about him. So he he shares some of the things that he's done to build AI into their systems and how they go about it. So here's the clip from last week's webinar that you can go watch of yourself if you care to. I'll leave the link in the show description. Walk me through some of the questions that you're asking as you evaluate. I I loved your one to five rating scale. I mean, I'd love to know
Parrish Gunnels: Yeah.
Samuel Hill: what what a one looks like in your rating scale. And then you know maybe those, you know, threes, fours, potentially even fives.
Parrish Gunnels: Yeah. Well, we we I we're getting to where I it's an ad hoc thing right now. So we're we are trying to move towards a a documented process and flow of how it works, right? But in general, we're trying to educate on the business side of things like, hey, because of the things we've experienced and the things we've seen, involve us early. Like I I don't wanna sit right, I don't wanna sit on every demo that of of every product that you're looking at, but If you're getting down to the one or two and you think this is gonna fit your need, then bring us in early. so we're and then we're asking for a meeting really with their engineers basically as if as much as we can. And say so and we're telling them up front, like we need to understand what is your AI? What are you you know, we're asking, what are you using, right? Are you using Claude? Are you using open AI are you using something else, right? We had an a company the other day said they're using Nova and I and I didn't know what that was. And I went and asked my counterpart and he's like, yeah, that's the stuff AWS has been building and it doesn't have a great reputation. I'm like, okay. So so we're trying to I'd say when we get on the on a call with them, we're we're asking those basic questions about how do you have your LM structured, how is it connected within the platform, etc. We're asking them point blank, did you code? Did you code that service in AI? If you did, can you can you give us a percentage of what you think that is, right? and that usually is followed up with are you do you have life cycle processes in place for development, right? So in other words, have you gone and checked that code? Have you scanned it? Have you you know what have you done?
Samuel Hill: Right. Right.
Parrish Gunnels: then we're asking for some sort of architectural diagram, like to show me how the service the front end platform, if it's client, connected or not, how all that is is geared together. and it's a lot of the QA is a lot of like depending on what how they answer it to, right? and you start getting a feeling of red flags or this seems fairly solid type thing. we also try to have we do have created a center of excellence in inside our organization where these sorts of ideas and things can be submitted because there is a a a an aspect where we can build on our own. and we have a a a great I think we're we're we have been down the journey of you know data centralization and that sort of stuff. So I think that
Samuel Hill: Mm-hmm. Mm-hmm.
Parrish Gunnels: our data lake and those sorts of things and how we handle is is good. And so that make gives me more confidence about building it internally just because it's our licensed AI or LLM that we've we've done. It is our data through our data lake. and the AI team that's doing all that development work is putting an orchestration layer together. There's controls and stuff. so there's just there's just more overall control in order to do that with, right? So there's still there's still a need, I think, for third party services, but I think with AI and some of the risk lending itself to Bringing things back in-house and building it is also a very high consideration, I think, that a lot of companies need to make. we definitely ask them those basic things, and then we also get into hallucinations, model validation, you know, all the quality pieces that, you know, what are you doing to make sure. quality of the response is good. The co what's the confidence level of those responses, right? So we're asking those and and you'll find there's a lot of companies that can't answer those questions, right? So that immediately goes, okay, well, you know, and and we we had one, I'll just share one story. We had one company we were talking with, and we were kind of in our heads putting together the timeline of when they started and where they are today, etc. And based on some other questions they had answered and then we realized that they had probably coded some of the service in OpenAI and then probably at some point use claud code. And then we made the comment to them that, well, okay, if you had coded in Claude code back past, you know, earlier than this date, right, you realize Anthropic came out and said there's problems, right? And you should go back and review and look. And they didn't know that. So that was scary.
Samuel Hill: Okay, a lot of interesting things happening in the cybersecurity market, if you will. There's a lot of the business of cybersecurity. I'm fascinated by it. Obviously, I live and breathe this every single day at work. but interesting in the data security space, we've seen a lot of consolidation. There's a lot of companies that have started and are working their way towards you know, viability, funding, all the things, offering solutions around data security on the wide spectrum of data security. I think we're seeing this across AI security techniques. I think we're seeing this across identity security techniques, things that are really innovating rapidly or being forced to innovate very rapidly in this AI world. And what we're seeing, so one reported and confirmed report is Proofpoint, one of the large DLP companies, very specific, very well known for their email DLP, well trusted company across many organizations, buying Veronis, a one of the early players of. Well-established players in the DSPM space. And this is a this is a very large acquisition, but it's like six billion plus dollars. That's billion with a B. A lot of money. Now, private equity is involved. I think Proof Point is owned by Toma Bravo and then pro and now Veronis, which was in the public market, I believe, is coming into a private equity acquisition. Who knows what's going to happen to these two organizations? the the the thesis, right, for this acquisition play is. That one plus one will equal three or potentially four, if you will, bringing together some DSPM capabilities into the proof point world, into the proof point offering. And also then adding more DLP capabilities, more data security, data in motion capabilities for the Veronist customers in the Veronus world. I don't know what will happen. I know a lot of people are going to be nervous if that's an established, trusted platform in your organization. Acquisitions are hard. It's one of the most difficult things to do in all of businesses. Working your way through an acquisition where what does the roadmap look like? What's the feature delivery look like? Now you have product teams and product organizations that are having to figure out, you know, what resources do we have, what funding do we have, what budget do we have, what's the direction overall? You know, fortunately, these are fairly distinct product lines that should retain a lot of the engineering talent and people to be able to keep moving them forward. So if that's a trusted thing for you, you know, who knows? We obviously pay close attention to it. I know I'll be keeping track of because that's a That's a large amount of money being exchanged to combine a couple of data security, established companies, again in the thesis of one powerhouse company that can really well secure data, both at rest across structured and unstructured data environments, and then in motion, as proof what went off, right? So who knows? hopefully it doesn't go the way of other acquisitions where the complications of all these teens and all this product and this backlog and The the roadmaps and all the stuff, the technical debt that both organizations inevitably hold because every organization has it. How does that all play out? So a lot of questions about that. The other one is CrowdStrike entering the DLP game, if you will, buying an emerging site. Well, sorry, let me back it up a little bit. It's reported that they're in the final stages of this acquisition. Who knows what what will actually end up happening, what the valuation of the acquisition is, what the actual numbers turn out to be. A lot of times though, where there is smoke, there is fire, and you know, CrowdStrike being on the endpoint already, well trusted across so many organizations to provide endpoint security, adding some DLP components, some more advanced DLP DLP components. They've offered that a little bit. They've had some offering in there, but but it's not really taken off. It's not really been a thing that a lot of customers have trusted for a lot of reasons. But being so well entrenched on the endpoint, that's probably the hardest thing to do in DLP is endpoint, the endpoint agent itself. And CrowdStrike obviously has that nailed. So buying a company like Orion for whatever the number ends up being, if it does end up being confirmed and going forward, we'll see. But the business of cybersecurity is just so fascinating because now how do you take this, young, fast-moving startup and incorporate it into the well-entrenched, well-trusted, risk-averse, if you if you believe that, CrowdStrike world. Even though they now have access to all of the endpoints that CrowdStrike has under management, which which totals probably in the billions a lot of endpoint agents out there now that they're gonna try and bring it all together. So, how do you mesh these two things? I don't know. I'm not paid enough to know all that, but I just think it's fascinating to see the consolidation because data security has become a hyper focused area, especially as AI is so rampantly running through organizations, it's touching all the data. Now it becomes a DLP problem, a data security problem of how do we keep data safe in this new AI world. So keep your eyes open. If you're customers in any of these places, obviously questions to ask of your vendors and of your partners, figure out, hey, what's the blast radius from this acquisition in my organization and what am I doing here? And then what's our strategy going into next year to either mitigate that blast radius, take advantage perhaps of some economies of scale or whatever those things are, but just new realities in the world of data security. I expect we'll see more. I do expect we'll see more consolidation as companies kind of try to figure out trying to get their hands on how do we secure the data that AI uses. Hey, it is Cybersecurity Awareness Month. And so all over security teams are finding ways to make their users aware of cybersecurity concerns. You know, nothing beats the front line where a user can spot the weird sounding written email or the, you know, know that this is not the right thing to do, being aware of best cybersecurity practices. And also, it's a great time of year to remember and celebrate and honor those women and men who are working so hard to keep data safe. And also for cybersecurity companies, their marketing teams also work extra hard to try and deliver things that are are engaging and fun for it. So, in that spirit, my marketing team is forcing me, hands tied behind my back, to watch videos that they've put together for Cybersecurity Awareness Month. So We're gonna watch them together. I'm gonna react to them and well, let's see what they've got in store for me. yeah, all the all the time. Okay. All right. Yeah, hey, I like that one. yeah, taking the time of user education. It's something that, everybody does all year and in in this month particularly. And how much you have to educate, especially in this AI world where AI is such a prevalent thing and users are doing all sorts of stuff. but I love the juxtaposition now. It's you get to do Oktoberfest instead of regular other stuff for October, the month of cybersecurity awareness month. So I thought that was kinda I thought that was kinda cute. All right, let's take a look at this second video. Yeah, all all the time. That happens. man, hey, I I like that one too. I mean, obviously something you educate users on is you don't share that file, that sensitive thing to everyone with a link, and yet they still do, and we haven't educate all the time. I'm not sure about the Yankees, man. Like I would imagine you're either you either love the Yankees or you really don't love the Yankees. And a lot of people fall into that. I don't love the Yankees camp. So but I guess there's a little, you know, silver lining. The Yankees got bounced from the playoffs. so they are no longer playing October baseball. So I I don't know. But either way, a a cute video. again, the reward of being able to watch your team in October versus having to do so much user education on stuff. Yeah, good job. All right, let's watch this third video together. Wait, there's more. Finally, one entire day without meetings that could have been slacking. that'd be a gift. With mine, you can automate many of the tedious data security. Projects include rested analysts, fewer single points of failure, and a CISO with newfound empathy. Donuts sold separately. Let's mind what matters. Alright, that's that's kinda that's kinda cute. I I mean I guess what would a team really want, you know, instead of a pizza party, you know, which a lot I no, nothing wrong with pizza, honoring your team with some some great lunch or food or things like that. But but yeah, hey, the the day without meetings, that could have been Slack messages. I mean, sign me up. That sounds absolutely amazing. it sounds really great, great video. That was I really like that one. All right, let's watch this last video together. And so the board presentation is not tagged as confidential and slips right past the tools meant to protect it. Looks like another successful user training. Mind can automatically apply sensitivity labels to your data so you no longer rely on users. Or do it for them. This October, you can admire the foliage instead of retraining people. Be aware this October, not anxious. man, that's the classic, right? The annual user training where you you they have the videos of the online content they have to go through and they they they pass the knowledge checks and a test for it and it's great, man. We gotta do it. gotta hit those marks for all the compliance reasons and everything. everybody everybody does it. The completion rate's really high. But were they actually watching it? Did they actually make a difference? And so I love gosh, I love Bob. We all know Bob, right? He's got definitely somebody we all know. And I love the ending where they know now I get to go look at foliage and be see the be outside and enjoy the fall time and and all that kind of stuff. That was pretty. I would love to go hiking there. It looks looks great. So great job. congrats, mind marketing. Those are great videos. I think they really summate the problem really well of of how hard it is to train users to keep them on the straight and narrow, if you will, in cybersecurity best practice and and some fun, cool videos as well that kind of help tell that story. Well, hey, all right, that's gonna do it for another episode of Mind, What Matters. I really appreciate you hanging out with us this week. And as always, if you could like and subscribe to the podcast so you can get notified of new episodes and really just helps us be found by other cybersecurity practitioners who want to maybe put a little more humanity in their day-to-day. Hey, with all that to be said, I'm really grateful for your time. My name is Samuel Hill, and that's all for now.
