Latest / Protected (by Corsair) / Are Your Employees Using AI Without You Knowing? The Hidden Cybersecurity Risks
Transcript
- 0:02[SPEAKER_01]: Welcome to Protected by Corsair USA.
- 0:05[SPEAKER_01]: This podcast is for business owners, executives, and professionals who want to understand cybersecurity and technology best practices.
- 0:13[SPEAKER_01]: We talk about real world security threats facing businesses today.
- 0:17[SPEAKER_01]: What you can do to protect your people, your data, your business, and how technology, including AI, is changing the cybersecurity landscape.
- 0:26[SPEAKER_01]: I'm Brian Gupers, partner in VP of Business Development for Corsair USA.
- 0:30[SPEAKER_00]: And I'm Rick Feldez, founder of Corsair USA, and today we're kicking off a new series, focused on artificial intelligence or AI.
- 0:40[SPEAKER_00]: Now we're not going to spend the series talking about building AI models much more practical.
- 0:45[SPEAKER_00]: What happens to your business when employees start incorporating AI into everyday tasks?
- 0:52[SPEAKER_01]: They'd be the even bigger question is what happens when they're already using it and nowhere knows about it.
- 0:57[SPEAKER_00]: Well, that's exactly where a lot of businesses find themselves today.
- 1:00[SPEAKER_00]: Leadership may not know what information is being shared by AI applications.
- 1:05[SPEAKER_00]: What are businesses actually using AI for right now?
- 1:08[SPEAKER_00]: Well, a few years ago, AI felt like something reserved for big technology companies, but not anymore.
- 1:15[SPEAKER_00]: Today, sales teams are using it to prepare for meetings.
- 1:18[SPEAKER_00]: Marketing teams are creating content with it.
- 1:21[SPEAKER_00]: Accountants are analyzing data.
- 1:27[SPEAKER_00]: attorneys are summarizing documents.
- 1:29[SPEAKER_00]: It's showing up everywhere and most employees don't view that as an IT decision You nailed it.
- 1:35[SPEAKER_00]: You're not thinking I'm deploying new software if they can I'm asking a question and that distinction is what creates a new cyber security challenge
- 1:44[SPEAKER_01]: let's make this real.
- 1:46[SPEAKER_01]: Let's say an employee is working on a proposal.
- 1:48[SPEAKER_01]: They copy and paste some information into chatGPT and ask it to make the writing sound more professional.
- 1:53[SPEAKER_01]: What's the risk?
- 1:54[SPEAKER_00]: It be there's no risk at all and that's what can complicate this scenario.
- 1:58[SPEAKER_00]: The issue isn't AI tool itself.
- 2:01[SPEAKER_00]: The issue is what information they put into it.
- 2:05[SPEAKER_00]: But maybe that's not a big deal.
- 2:07[SPEAKER_00]: But if it's a customer contract, pricing information, employee records, confidential financial information or client information, now we're having a very different conversation.
- 2:18[SPEAKER_01]: Most employees are trying to do anything wrong.
- 2:20[SPEAKER_01]: They're simply trying to be more productive and stay on top of current technology.
- 2:25[SPEAKER_00]: It's right, they found it too old, it saves time, and might save them a couple of hours of work, so that's smart.
- 2:32[SPEAKER_00]: But from the company's perspective, they may have just shared sensitive business information outside normal security controls.
- 2:39[SPEAKER_00]: And that's where we're here to tell you, shadow AI.
- 2:42[SPEAKER_00]: They gave Shadow AI as the next evolution of Shadow IT.
- 2:47[SPEAKER_00]: Years ago, employees would sign up for cloud services or install applications without IT approval.
- 2:54[SPEAKER_00]: And today, they could create an AI account in seconds and start using it immediately.
- 2:59[SPEAKER_00]: No software installation.
- 3:01[SPEAKER_00]: They simply use a browser, no procurement process, and no security review.
- 3:07[SPEAKER_00]: And no one knows it's happening.
- 3:08[SPEAKER_00]: And that's what makes it difficult to manage.
- 3:11[SPEAKER_00]: And here's the reality.
- 3:13[SPEAKER_00]: Most organizations aren't going to solve it by simply banning AI.
- 3:18[SPEAKER_00]: So what's the solution?
- 3:19[SPEAKER_00]: People are going to use it anyway.
- 3:20[SPEAKER_00]: Well that's right.
- 3:21[SPEAKER_00]: The goal shouldn't be to stop people from using AI.
- 3:24[SPEAKER_00]: The goal should be helping them use AI safely.
- 3:27[SPEAKER_01]: So let's talk about the tools people hear about every day, Chad GPT, Cole Pilate, Gemini, Claude, and the security perspective.
- 3:36[SPEAKER_01]: Aren't they all basically the same?
- 3:38[SPEAKER_00]: No, no, there's a huge difference between consumer AI tools.
- 3:43[SPEAKER_00]: an enterprise AI platform.
- 3:45[SPEAKER_00]: The permissions, security controls, account configurations, administrative oversight, and data protections can vary significantly from one platform to the other.
- 3:58[SPEAKER_00]: So the takeaway is in AI as dangerous.
- 4:00[SPEAKER_00]: No, the takeaway is that organizations should not assume that every AI implementation
- 4:10[SPEAKER_00]: There's a big difference between an employee using a personal AI account and a company deploying an enterprise platform that has been properly configured and managed.
- 4:21[SPEAKER_00]: So, if co-pilot is part of Microsoft, aren't we secure?
- 4:25[SPEAKER_00]: Microsoft has invested heavily in security.
- 4:27[SPEAKER_00]: That's true.
- 4:28[SPEAKER_00]: But security isn't just about the product.
- 4:31[SPEAKER_00]: It's also your environment.
- 4:33[SPEAKER_00]: If your permissions are poorly configured and managed, AI doesn't magically fix that.
- 4:38[SPEAKER_00]: If employees already have the information that they need to have, AI can make that information easier to find, summarize, and use it.
- 4:48[SPEAKER_00]: So AI is increasing the trouble and it's exposing.
- 4:51[SPEAKER_00]: Yes, sometimes AI acts like a spotlight on security issues that already existed.
- 4:56[SPEAKER_00]: So Rick, what's the first question business owner should be asking right now?
- 5:00[SPEAKER_00]: Well, in most cases, the first slide defense are the employees.
- 5:05[SPEAKER_00]: So having an AI security policy in place is where I would start.
- 5:09[SPEAKER_00]: They know it would provide training to the employees to let them know what type of information should not be being accessed by AI.
- 5:18[SPEAKER_00]: For example, customer data, employee information, contracts, financial records, source code, business plans, intellectual property, anything confidential.
- 5:29[SPEAKER_00]: People don't always recognize sense of information when they're working with it.
- 5:32[SPEAKER_00]: Well, someone might say, it's just to spreadsheet.
- 5:35[SPEAKER_00]: Well, what's inside it?
- 5:36[SPEAKER_00]: It's just a customer list.
- 5:38[SPEAKER_00]: You have to look beyond the file and focus on the actual data.
- 5:42[SPEAKER_01]: So let's walk through a real world scenario.
- 5:44[SPEAKER_01]: A customer sends a complaint and employee copies the email into chat GPT and says right at professional response.
- 5:51[SPEAKER_00]: Okay, so the concern is that the customer information may have been entered into a system.
- 5:58[SPEAKER_00]: The company has not reviewed, approved, or configured.
- 6:01[SPEAKER_00]: The employee may also have no idea what contractual regulatory or privacy obligation applies to that information.
- 6:09[SPEAKER_00]: So what you do organizations do.
- 6:11[SPEAKER_00]: It's all about implementation of an AI policy and employee training.
- 6:16[SPEAKER_00]: And AI policy can be simple.
- 6:19[SPEAKER_00]: Use approved AI tools.
- 6:21[SPEAKER_00]: Don't enter confidential information.
- 6:24[SPEAKER_00]: and provide employees with approved workflows that they can follow.
- 6:28[SPEAKER_01]: So we're not saying don't use AI.
- 6:30[SPEAKER_01]: We're saying use AI responsibly.
- 6:32[SPEAKER_01]: And here's something many businesses may not realize.
- 6:35[SPEAKER_01]: AI isn't just chatGPT and it's built into email platforms.
- 6:40[SPEAKER_01]: CRM systems, accounting software, HR tools.
- 6:43[SPEAKER_01]: customer service platforms, in productivity speeds.
- 6:46[SPEAKER_00]: And that's one of the biggest shifts happening right now.
- 6:49[SPEAKER_00]: You don't have to look for AI anymore.
- 6:52[SPEAKER_00]: AI is coming to you.
- 6:54[SPEAKER_01]: You open your email and there's AI.
- 6:56[SPEAKER_01]: You open your CRM and there's AI.
- 6:58[SPEAKER_01]: You open a document and there's AI.
- 7:00[SPEAKER_00]: And that can be a good thing.
- 7:02[SPEAKER_00]: Because those tools can operate inside existing security and compliance frameworks, but configuration still matters.
- 7:10[SPEAKER_00]: So let's wrap up with one thing
- 7:14[SPEAKER_00]: Okay, here it is.
- 7:16[SPEAKER_00]: AI is not just another application.
- 7:18[SPEAKER_00]: It's changing how employees interact with company information.
- 7:22[SPEAKER_00]: Consequently, cybersecurity has to change too.
- 7:25[SPEAKER_00]: So don't just ask, I might system secure.
- 7:27[SPEAKER_00]: That's right.
- 7:28[SPEAKER_00]: Ask, do we have an AI policy in place?
- 7:32[SPEAKER_00]: Have our employees been trained on the AI policy?
- 7:36[SPEAKER_00]: What AI tools are employees using?
- 7:39[SPEAKER_00]: What systems
- 7:43[SPEAKER_00]: what actions can those systems perform?
- 7:47[SPEAKER_00]: This is basically another level of risk assessment.
- 7:50[SPEAKER_01]: My biggest takeaway is that AI is creating an entirely new cybersecurity problem.
- 7:55[SPEAKER_01]: It's changing the way existing cybersecurity challenges show up.
- 7:59[SPEAKER_00]: And we're only getting started.
- 8:01[SPEAKER_00]: In the next episode, we'll dive into AI agents and what can happen when AI can actually take actions on your behalf.
- 8:09[SPEAKER_00]: That's where things really get interesting.
- 8:11[SPEAKER_00]: and potentially more complex, especially from a security standpoint.
- 8:16[SPEAKER_00]: You've been listening to Protected by Corsair on Rick Valdez.
- 8:20[SPEAKER_01]: I'm Brian Gilchrist.
- 8:21[SPEAKER_01]: If you're not sure where your business stands today, start with the cybersecurity risk assessment.
- 8:25[SPEAKER_01]: Visit us on coursener cms.com.
- 8:28[SPEAKER_01]: If you found this episode helpful, subscribe to the podcast and share it with another business.
- 8:32[SPEAKER_01]: Thanks for listening to Protected by Corsair USA.
- 8:35[SPEAKER_01]: See you next time.
- 8:36[SPEAKER_01]: Thanks for listening.